Language Not Auditory Experience is Related to Parent-Reported Executive Functioning in Preschool-Aged Deaf and Hard of Hearing Children

Load Packages

Load the following libraries. If they are not installed, run install.packages (“packagename”)

#from https://www.vikram-baliga.com/blog/2015/7/19/a-hassle-free-way-to-verify-that-r-packages-are-installed-and-loaded
packages = c("tidyverse","ggplot2","beeswarm","colorspace", "stargazer", "qwraps2", "gridExtra","ggpubr","car","olsrr","ggbeeswarm", "broom", "rmarkdown", "table1", "kableExtra", "epitools", "FSA", "Hmisc", "ufs")
package.check <- lapply(packages, FUN = function(x) {
  if (!require(x, character.only = TRUE)) {
    install.packages(x, dependencies = TRUE)
    library(x, character.only = TRUE)
  }
})

#verify they are loaded
search()

Setup Primary Dataframe

First, import the data from the data (CSV) file

#setwd("~/etc") #This should be wherever your file is saved
BRIEF <- read.csv("SLaM_BRIEF-P raw data 20200911.csv", na.strings = c("N/A", "", "Unknown", "Excluded"))

Then, subset imported dataframe to remove participants missing age or SES information. We also removed deaf participants who had unilateral hearing loss.

This results in 123 participants

Add 4 columns:

Language Modality
Language Timing
4-group split
3-way group split

BRIEF <- subset(BRIEF, BRIEF$AgeMonths < 100 & BRIEF$AgeMonths!="" & BRIEF$SES..3.66.!="" & BRIEF$UnilateralHearingLoss. == "No")

BRIEF$Language_Modality <- factor(ifelse(BRIEF$Group_4cat == "English Early" | BRIEF$Group_4cat == "English Later", "English", "ASL"), levels = c("English", "ASL"))
BRIEF$Language_Timing <- factor(as.character(BRIEF$Group_2cat), levels = c("Early", "Later"), exclude="")

BRIEF$LanguageGroup <- as.factor(factor(as.character(BRIEF$Group_4cat), levels = c("English Early", "ASL Early", "English Later", "ASL Later"), labels = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), exclude=NA))

BRIEF$Hearing_Level_3_Cat <- as.factor(factor(as.character(BRIEF$Hearing_Level_3_Cat), levels = c("TH", "deaf", "HoH"), labels = c("Typically Hearing", "Deaf", "Hard of Hearing"), exclude=NA))

BRIEF$Hearing.Device <- dplyr::recode(BRIEF$Hearing.Device, "BAHA" = "Other")
BRIEF$Hearing.Device <- as.factor(factor(as.character(BRIEF$Hearing.Device), levels = c("HA", "CI", "HA & CI", "Other", "None"), labels = c("Hearing Aid", "Cochlear Implant", "Hearing Aid & Cochlear Implant", "Other", "None"), exclude=NA))

#Recode Race to have fewer categories
BRIEF$Race_recoded <- dplyr::recode(as.character(BRIEF$Race), 'Asian' = "Asian", 'Black or African American'="Black or African American", 'More than One'="More than one", 'Other'="Other/Missing", 'Unsure, or prefer not to answer' = "Other/Missing", 'White'="White", .missing="Other/Missing")

#update ORDER of groups
BRIEF$Race_recoded <- factor(BRIEF$Race_recoded, levels = c("Asian", "Black or African American", "White", "More than one", "Other/Missing"), labels = c("Asian", "Black or African American", "White", "More than one", "Other/Missing"))

#Recode Ethnicity to have fewer categories
BRIEF$Ethnicity_recoded <- dplyr::recode(as.character(BRIEF$Ethnicity), 'Hispanic' = "Hispanic", 'NonHispanic'="Non-Hispanic", 'PreferNotToAnswer'="Prefer not to answer", .missing="Missing")

#update ORDER of groups
BRIEF$Ethnicity_recoded <-factor(BRIEF$Ethnicity_recoded, levels = c("Hispanic", "Non-Hispanic", "Prefer not to answer", "Missing"), labels = c("Hispanic", "Non-Hispanic", "Prefer not to answer", "Missing"))

Demographic Information

Demographics for the Entire Sample

table1::label(BRIEF$AgeMonths) <- "Age (Months)"
table1::label(BRIEF$SES..3.66.) <- "SES"
table1::label(BRIEF$Sex) <- "Sex"
table1::label(BRIEF$Race_recoded) <- "Race"
table1::label(BRIEF$Ethnicity_recoded) <- "Ethnicity"
 
table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded, data = BRIEF)

	Overall (N=123)
Age (Months)
Mean (SD)	60.1 (13.5)
Median [Min, Max]	59.0 [37.0, 91.0]
SES
Mean (SD)	49.8 (14.7)
Median [Min, Max]	54.0 [8.00, 66.0]
Sex
Female	66 (53.7%)
Male	57 (46.3%)
Race
Asian	3 (2.4%)
Black or African American	1 (0.8%)
White	104 (84.6%)
More than one	10 (8.1%)
Other/Missing	5 (4.1%)
Ethnicity
Hispanic	10 (8.1%)
Non-Hispanic	97 (78.9%)
Prefer not to answer	1 (0.8%)
Missing	15 (12.2%)

Table 1.

table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded  | LanguageGroup, data = BRIEF, overall=F)

	Typically Hearing (N=46)	Early ASL (N=26)	Later English (N=23)	Later ASL (N=28)
Age (Months)
Mean (SD)	54.7 (10.9)	60.8 (14.4)	60.7 (10.9)	68.0 (14.6)
Median [Min, Max]	54.5 [37.0, 85.0]	57.0 [41.0, 91.0]	60.0 [37.0, 78.0]	71.5 [37.0, 90.0]
SES
Mean (SD)	55.6 (8.83)	48.2 (16.8)	48.8 (14.1)	42.7 (17.4)
Median [Min, Max]	56.0 [21.5, 66.0]	56.5 [11.0, 66.0]	51.0 [8.00, 66.0]	49.0 [9.00, 62.0]
Sex
Female	24 (52.2%)	16 (61.5%)	13 (56.5%)	13 (46.4%)
Male	22 (47.8%)	10 (38.5%)	10 (43.5%)	15 (53.6%)
Race
Asian	0 (0%)	0 (0%)	0 (0%)	3 (10.7%)
Black or African American	0 (0%)	0 (0%)	1 (4.3%)	0 (0%)
White	43 (93.5%)	24 (92.3%)	18 (78.3%)	19 (67.9%)
More than one	3 (6.5%)	1 (3.8%)	3 (13.0%)	3 (10.7%)
Other/Missing	0 (0%)	1 (3.8%)	1 (4.3%)	3 (10.7%)
Ethnicity
Hispanic	3 (6.5%)	0 (0%)	2 (8.7%)	5 (17.9%)
Non-Hispanic	42 (91.3%)	18 (69.2%)	17 (73.9%)	20 (71.4%)
Prefer not to answer	0 (0%)	1 (3.8%)	0 (0%)	0 (0%)
Missing	1 (2.2%)	7 (26.9%)	4 (17.4%)	3 (10.7%)

Comparing SES across Language Groups

ggboxplot(BRIEF, x = "LanguageGroup", y = "SES..3.66.", 
          color = "LanguageGroup",
          order = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"),
          ylab = "SES", xlab = "Language Group")

ggplot(data=BRIEF, mapping=aes(x=SES..3.66.))+ geom_histogram(binwidth=10) + facet_grid(~LanguageGroup)

leveneTest(SES..3.66.~LanguageGroup, data=BRIEF)

## Levene's Test for Homogeneity of Variance (center = median)
##        Df F value  Pr(>F)  
## group   3  3.6537 0.01457 *
##       119                  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

SES_LangGrp <- kruskal.test(SES..3.66.~LanguageGroup, data=BRIEF)
SES_LangGrp

## 
##  Kruskal-Wallis rank sum test
## 
## data:  SES..3.66. by LanguageGroup
## Kruskal-Wallis chi-squared = 11.705, df = 3, p-value = 0.008465

dunnTest(SES..3.66.~LanguageGroup, data=BRIEF, method="hochberg")

## Dunn (1964) Kruskal-Wallis multiple comparison

##   p-values adjusted with the Hochberg method.

##                          Comparison         Z      P.unadj       P.adj
## 1             Early ASL - Later ASL  1.712379 0.0868268640 0.347307456
## 2         Early ASL - Later English  0.616999 0.5372354018 0.537235402
## 3         Later ASL - Later English -1.029644 0.3031773783 0.606354757
## 4     Early ASL - Typically Hearing -1.334535 0.1820284853 0.546085456
## 5     Later ASL - Typically Hearing -3.311757 0.0009271197 0.005562718
## 6 Later English - Typically Hearing -1.973772 0.0484076294 0.242038147

#selected hochberg correction method referencing this site: https://towardsdatascience.com/an-overview-of-methods-to-address-the-multiple-comparison-problem-310427b3ba92

Findings:

Later ASL has significantly lower SES than Typically Hearing
No other between-group differences in SES

Comparing Age across Language Groups

ggboxplot(BRIEF, x = "LanguageGroup", y = "AgeMonths", 
          color = "LanguageGroup",
          order = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"),
          ylab = "Age (Months)", xlab = "Language Group")

ggplot(data=BRIEF, mapping=aes(x=AgeMonths))+ geom_histogram(binwidth=6) + facet_grid(~LanguageGroup)

leveneTest(AgeMonths~LanguageGroup, data=BRIEF)

## Levene's Test for Homogeneity of Variance (center = median)
##        Df F value Pr(>F)
## group   3   1.255 0.2931
##       119

Age_LangGrp <- kruskal.test(AgeMonths~LanguageGroup, data=BRIEF)
Age_LangGrp

## 
##  Kruskal-Wallis rank sum test
## 
## data:  AgeMonths by LanguageGroup
## Kruskal-Wallis chi-squared = 16.981, df = 3, p-value = 0.000713

dunnTest(AgeMonths~LanguageGroup, data=BRIEF, method="hochberg")

## Dunn (1964) Kruskal-Wallis multiple comparison

##   p-values adjusted with the Hochberg method.

##                          Comparison          Z      P.unadj        P.adj
## 1             Early ASL - Later ASL -2.0851898 3.705207e-02 0.1852603675
## 2         Early ASL - Later English -0.3394709 7.342550e-01 0.7342550228
## 3         Later ASL - Later English  1.6727536 9.437580e-02 0.2831274122
## 4     Early ASL - Typically Hearing  1.6716134 9.460059e-02 0.1892011775
## 5     Later ASL - Typically Hearing  4.0804060 4.495711e-05 0.0002697427
## 6 Later English - Typically Hearing  1.9865461 4.697272e-02 0.1878908797

Findings:

Later ASL group significantly older than Typically Hearing group
No other between-group differences in Age

Results

Question 1: Do DHH children with access to sign language but not auditory input from birth exhibit executive functioning skills comparable to their typically-hearing peers?

Approach: Compare (via Welch’s t-tests) the BRIEF raw scores of Early English group to those of Early ASL group

Create dataframe with only children exposed to language early and perform t-tests for all BRIEF-P scales

BRIEF_early <- subset(BRIEF, BRIEF$Language_Timing=="Early")

Demographics for the Early-exposed subset

table1::label(BRIEF_early$AgeMonths) <- "Age (Months)"
table1::label(BRIEF_early$SES..3.66.) <- "SES"
table1::label(BRIEF_early$Sex) <- "Sex"
table1::label(BRIEF_early$Race_recoded) <- "Race"
table1::label(BRIEF_early$Ethnicity_recoded) <- "Ethnicity"
 
table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded, data = BRIEF_early)

	Overall (N=72)
Age (Months)
Mean (SD)	56.9 (12.5)
Median [Min, Max]	55.0 [37.0, 91.0]
SES
Mean (SD)	52.9 (12.7)
Median [Min, Max]	56.0 [11.0, 66.0]
Sex
Female	40 (55.6%)
Male	32 (44.4%)
Race
Asian	0 (0%)
Black or African American	0 (0%)
White	67 (93.1%)
More than one	4 (5.6%)
Other/Missing	1 (1.4%)
Ethnicity
Hispanic	3 (4.2%)
Non-Hispanic	60 (83.3%)
Prefer not to answer	1 (1.4%)
Missing	8 (11.1%)

Comparing demographics by Language Modality (within early-exposed subset)

table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded  | LanguageGroup, data = BRIEF_early, overall=F)

	Typically Hearing (N=46)	Early ASL (N=26)
Age (Months)
Mean (SD)	54.7 (10.9)	60.8 (14.4)
Median [Min, Max]	54.5 [37.0, 85.0]	57.0 [41.0, 91.0]
SES
Mean (SD)	55.6 (8.83)	48.2 (16.8)
Median [Min, Max]	56.0 [21.5, 66.0]	56.5 [11.0, 66.0]
Sex
Female	24 (52.2%)	16 (61.5%)
Male	22 (47.8%)	10 (38.5%)
Race
Asian	0 (0%)	0 (0%)
Black or African American	0 (0%)	0 (0%)
White	43 (93.5%)	24 (92.3%)
More than one	3 (6.5%)	1 (3.8%)
Other/Missing	0 (0%)	1 (3.8%)
Ethnicity
Hispanic	3 (6.5%)	0 (0%)
Non-Hispanic	42 (91.3%)	18 (69.2%)
Prefer not to answer	0 (0%)	1 (3.8%)
Missing	1 (2.2%)	7 (26.9%)

Table 3.

BRIEF Scores & Welch two sample t-tests for two “Early” groups

table1::label(BRIEF_early$GEC_RawScore) <- "Global Executive Composite"
table1::label(BRIEF_early$Inhibit_RawScore) <- "Inhibition"
table1::label(BRIEF_early$Shift_RawScore) <- "Shift"
table1::label(BRIEF_early$Emotional.Control_RawScore) <- "Emotional Control"
table1::label(BRIEF_early$Working.Memory_RawScore) <- "Working Memory"
table1::label(BRIEF_early$Plan.Organize_RawScore) <- "Plan/Organize"

table1(~GEC_RawScore + Inhibit_RawScore + Shift_RawScore + Emotional.Control_RawScore + Working.Memory_RawScore + Plan.Organize_RawScore | Language_Modality, data = BRIEF_early, overall=F)

	English (N=46)	ASL (N=26)
Global Executive Composite
Mean (SD)	87.8 (18.7)	89.5 (16.1)
Median [Min, Max]	87.5 [63.0, 155]	90.0 [63.0, 132]
Inhibition
Mean (SD)	22.8 (5.93)	24.0 (6.06)
Median [Min, Max]	23.0 [16.0, 46.0]	22.5 [16.0, 37.0]
Shift
Mean (SD)	13.0 (3.11)	13.5 (3.13)
Median [Min, Max]	12.0 [10.0, 24.0]	12.5 [10.0, 21.0]
Emotional Control
Mean (SD)	15.0 (3.67)	14.7 (3.04)
Median [Min, Max]	14.0 [10.0, 24.0]	14.0 [10.0, 21.0]
Working Memory
Mean (SD)	22.7 (6.61)	23.1 (4.80)
Median [Min, Max]	20.5 [17.0, 48.0]	22.0 [17.0, 37.0]
Plan/Organize
Mean (SD)	14.3 (3.40)	14.2 (3.03)
Median [Min, Max]	13.5 [10.0, 23.0]	14.0 [10.0, 20.0]

t.test(BRIEF_early$GEC_RawScore~BRIEF_early$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early$GEC_RawScore by BRIEF_early$Language_Modality
## t = -0.40065, df = 58.665, p-value = 0.6901
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -10.065079   6.707219
## sample estimates:
## mean in group English     mean in group ASL 
##              87.78261              89.46154

t.test(BRIEF_early$Inhibit_RawScore~BRIEF_early$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early$Inhibit_RawScore by BRIEF_early$Language_Modality
## t = -0.81335, df = 51.079, p-value = 0.4198
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -4.164164  1.762826
## sample estimates:
## mean in group English     mean in group ASL 
##              22.76087              23.96154

t.test(BRIEF_early$Shift_RawScore~BRIEF_early$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early$Shift_RawScore by BRIEF_early$Language_Modality
## t = -0.57441, df = 51.763, p-value = 0.5682
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.976368  1.096769
## sample estimates:
## mean in group English     mean in group ASL 
##              13.02174              13.46154

t.test(BRIEF_early$Emotional.Control_RawScore~BRIEF_early$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early$Emotional.Control_RawScore by BRIEF_early$Language_Modality
## t = 0.30726, df = 60.411, p-value = 0.7597
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.363471  1.858454
## sample estimates:
## mean in group English     mean in group ASL 
##              14.97826              14.73077

t.test(BRIEF_early$Working.Memory_RawScore~BRIEF_early$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early$Working.Memory_RawScore by BRIEF_early$Language_Modality
## t = -0.29375, df = 65.462, p-value = 0.7699
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -3.103459  2.307473
## sample estimates:
## mean in group English     mean in group ASL 
##              22.71739              23.11538

t.test(BRIEF_early$Plan.Organize_RawScore~BRIEF_early$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early$Plan.Organize_RawScore by BRIEF_early$Language_Modality
## t = 0.14405, df = 57.092, p-value = 0.886
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.445442  1.669522
## sample estimates:
## mean in group English     mean in group ASL 
##              14.30435              14.19231

Findings: Early ASL group did not significantly differ from Typically Hearing group on any subscale ***

Question 2: Do age of access to auditory input and age of language exposure significantly predict executive functioning?

Approach: Linear regression models for the BRIEF-P raw scores on general composite score and each subscale predicted by age of auditory and age of language exposure, and other demographic characteristics (SES, sex, age)

Also testing whether other demographic variables or ways of characterizing group differences (Length of Auditory Exposure, Language Modality) are better predictive of BRIEF scores

Create dataframe for participants for whom we have Age of Auditory Exposure and Age of Language Exposure information

BRIEF <- dplyr::mutate(BRIEF, AoAE = ifelse(BRIEF$Age.of.Auditory.Exposure..mo.=="None", as.character(BRIEF$AgeMonths), as.character(BRIEF$Age.of.Auditory.Exposure..mo.)))

BRIEF$AoAE <- as.integer(BRIEF$AoAE)

BRIEF_AgeOf <- subset(BRIEF, BRIEF$AoAE!='' & BRIEF$Age.of.Language.Exposure..mo.!='')

This is the dataframe used for linear models in Question 2

Create numerical “AoAE” variable from Age of Auditory Exposure variable, converting “None” to current age
Change structure of Age of Auditory Exposure from character to integer
Remove participants for whom either of these two variables is missing
This results in 109 participants

Demographics for the Question 2 subset

table1::label(BRIEF_AgeOf$AgeMonths) <- "Age (Months)"
table1::label(BRIEF_AgeOf$SES..3.66.) <- "SES"
table1::label(BRIEF_AgeOf$Sex) <- "Sex"
table1::label(BRIEF_AgeOf$Race_recoded) <- "Race"
table1::label(BRIEF_AgeOf$Ethnicity_recoded) <- "Ethnicity"
table1::label(BRIEF_AgeOf$Age.of.Language.Exposure..mo.) <- "Age of Language Exposure (Months)"
table1::label(BRIEF_AgeOf$AoAE) <- "Age of Auditory Exposure (Months)"

 
table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded + Age.of.Language.Exposure..mo. + AoAE, data = BRIEF_AgeOf)

	Overall (N=109)
Age (Months)
Mean (SD)	59.2 (13.0)
Median [Min, Max]	59.0 [37.0, 91.0]
SES
Mean (SD)	51.2 (13.5)
Median [Min, Max]	54.5 [8.00, 66.0]
Sex
Female	58 (53.2%)
Male	51 (46.8%)
Race
Asian	3 (2.8%)
Black or African American	1 (0.9%)
White	94 (86.2%)
More than one	8 (7.3%)
Other/Missing	3 (2.8%)
Ethnicity
Hispanic	6 (5.5%)
Non-Hispanic	90 (82.6%)
Prefer not to answer	1 (0.9%)
Missing	12 (11.0%)
Age of Language Exposure (Months)
Mean (SD)	12.7 (18.8)
Median [Min, Max]	0 [0, 76.0]
Age of Auditory Exposure (Months)
Mean (SD)	19.9 (25.2)
Median [Min, Max]	4.00 [0, 91.0]

Comparing demographics by Language group (within Question 2 subset)

table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded + Age.of.Language.Exposure..mo. + AoAE | LanguageGroup, data = BRIEF_AgeOf, overall=F)

	Typically Hearing (N=46)	Early ASL (N=18)	Later English (N=23)	Later ASL (N=22)
Age (Months)
Mean (SD)	54.7 (10.9)	59.2 (14.5)	60.7 (10.9)	67.3 (14.0)
Median [Min, Max]	54.5 [37.0, 85.0]	57.0 [41.0, 91.0]	60.0 [37.0, 78.0]	71.0 [37.0, 90.0]
SES
Mean (SD)	55.6 (8.83)	49.0 (16.2)	48.8 (14.1)	46.0 (16.2)
Median [Min, Max]	56.0 [21.5, 66.0]	57.5 [19.0, 66.0]	51.0 [8.00, 66.0]	51.5 [14.0, 62.0]
Sex
Female	24 (52.2%)	9 (50.0%)	13 (56.5%)	12 (54.5%)
Male	22 (47.8%)	9 (50.0%)	10 (43.5%)	10 (45.5%)
Race
Asian	0 (0%)	0 (0%)	0 (0%)	3 (13.6%)
Black or African American	0 (0%)	0 (0%)	1 (4.3%)	0 (0%)
White	43 (93.5%)	17 (94.4%)	18 (78.3%)	16 (72.7%)
More than one	3 (6.5%)	0 (0%)	3 (13.0%)	2 (9.1%)
Other/Missing	0 (0%)	1 (5.6%)	1 (4.3%)	1 (4.5%)
Ethnicity
Hispanic	3 (6.5%)	0 (0%)	2 (8.7%)	1 (4.5%)
Non-Hispanic	42 (91.3%)	13 (72.2%)	17 (73.9%)	18 (81.8%)
Prefer not to answer	0 (0%)	1 (5.6%)	0 (0%)	0 (0%)
Missing	1 (2.2%)	4 (22.2%)	4 (17.4%)	3 (13.6%)
Age of Language Exposure (Months)
Mean (SD)	0 (0)	0 (0)	21.2 (16.2)	40.5 (12.5)
Median [Min, Max]	0 [0, 0]	0 [0, 0]	16.0 [0.500, 58.0]	36.5 [18.0, 76.0]
Age of Auditory Exposure (Months)
Mean (SD)	0 (0)	49.0 (24.9)	21.2 (16.3)	36.3 (24.5)
Median [Min, Max]	0 [0, 0]	52.5 [1.00, 91.0]	16.0 [0, 58.0]	35.0 [3.00, 79.0]

Table 2.

table1::label(BRIEF$Hearing_Level_3_Cat) <- "Pre-device hearing level (3 Category)"
table1::label(BRIEF$Hearing.Device) <- "Type of Hearing Device"
table1::label(BRIEF$AoAE) <- "Age of First Auditory Exposure (Months)"
table1::label(BRIEF$Age.of.Language.Exposure..mo.) <- "Age of First Language Exposure (Months)"

table1::table1(~Hearing_Level_3_Cat + Hearing.Device + AoAE + Age.of.Language.Exposure..mo.| LanguageGroup, data = BRIEF, overall=F)

	Typically Hearing (N=46)	Early ASL (N=26)	Later English (N=23)	Later ASL (N=28)
Pre-device hearing level (3 Category)
Typically Hearing	46 (100%)	0 (0%)	0 (0%)	0 (0%)
Deaf	0 (0%)	15 (57.7%)	13 (56.5%)	21 (75.0%)
Hard of Hearing	0 (0%)	7 (26.9%)	9 (39.1%)	5 (17.9%)
Missing	0 (0%)	4 (15.4%)	1 (4.3%)	2 (7.1%)
Type of Hearing Device
Hearing Aid	0 (0%)	11 (42.3%)	11 (47.8%)	12 (42.9%)
Cochlear Implant	0 (0%)	0 (0%)	9 (39.1%)	8 (28.6%)
Hearing Aid & Cochlear Implant	0 (0%)	0 (0%)	2 (8.7%)	1 (3.6%)
Other	0 (0%)	1 (3.8%)	1 (4.3%)	0 (0%)
None	46 (100%)	14 (53.8%)	0 (0%)	7 (25.0%)
Age of First Auditory Exposure (Months)
Mean (SD)	0 (0)	49.0 (24.9)	21.2 (16.3)	38.6 (26.5)
Median [Min, Max]	0 [0, 0]	52.5 [1.00, 91.0]	16.0 [0, 58.0]	36.0 [3.00, 90.0]
Missing	0 (0%)	8 (30.8%)	0 (0%)	5 (17.9%)
Age of First Language Exposure (Months)
Mean (SD)	0 (0)	0 (0)	21.2 (16.2)	42.0 (13.2)
Median [Min, Max]	0 [0, 0]	0 [0, 0]	16.0 [0.500, 58.0]	36.5 [18.0, 76.0]
Missing	0 (0%)	0 (0%)	0 (0%)	2 (7.1%)

Comparing Age of Auditory Exposure across Language Groups

ggboxplot(BRIEF, x = "LanguageGroup", y = "AoAE", 
          color = "LanguageGroup",
          order = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"),
          ylab = "Age of Auditory Exposure (Months)", xlab = "Language Group")

## Warning: Removed 13 rows containing non-finite values (stat_boxplot).

ggplot(data=BRIEF, mapping=aes(x=AoAE))+ geom_histogram(binwidth=15) + facet_grid(~LanguageGroup)

## Warning: Removed 13 rows containing non-finite values (stat_bin).

leveneTest(AoAE~LanguageGroup, data=BRIEF)

## Levene's Test for Homogeneity of Variance (center = median)
##        Df F value    Pr(>F)    
## group   3  26.141 9.781e-13 ***
##       106                      
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

AoAE_LangGrp <- kruskal.test(AoAE~LanguageGroup, data=BRIEF)
AoAE_LangGrp

## 
##  Kruskal-Wallis rank sum test
## 
## data:  AoAE by LanguageGroup
## Kruskal-Wallis chi-squared = 88.967, df = 3, p-value < 2.2e-16

dunnTest(AoAE~LanguageGroup, data=BRIEF, method="hochberg")

## Warning: Some rows deleted from 'x' and 'g' because missing data.

## Dunn (1964) Kruskal-Wallis multiple comparison

##   p-values adjusted with the Hochberg method.

##                          Comparison         Z      P.unadj        P.adj
## 1             Early ASL - Later ASL 0.8923039 3.722301e-01 3.722301e-01
## 2         Early ASL - Later English 2.2593334 2.386265e-02 7.158796e-02
## 3         Later ASL - Later English 1.4588767 1.445991e-01 2.891981e-01
## 4     Early ASL - Typically Hearing 7.6412506 2.151219e-14 1.290732e-13
## 5     Later ASL - Typically Hearing 7.2191701 5.230578e-13 2.615289e-12
## 6 Later English - Typically Hearing 5.5346044 3.119315e-08 1.247726e-07

Findings:

Typically hearing group has significantly earlier age of auditory exposure than all three DHH groups (as expected)
Later English group did not differ in terms of age of auditory exposure compared to both Early and Later ASL groups
Early and Later ASL groups did not differ in terms of their age of auditory exposure

Comparing Age of Language Exposure across Language Groups

ggboxplot(BRIEF, x = "LanguageGroup", y = "Age.of.Language.Exposure..mo.", 
          color = "LanguageGroup",
          order = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"),
          ylab = "Age of Language Exposure (Months)", xlab = "Language Group")

## Warning: Removed 2 rows containing non-finite values (stat_boxplot).

ggplot(data=BRIEF, mapping=aes(x=Age.of.Language.Exposure..mo.))+ geom_histogram(binwidth=10) + facet_grid(~LanguageGroup)

## Warning: Removed 2 rows containing non-finite values (stat_bin).

leveneTest(Age.of.Language.Exposure..mo.~LanguageGroup, data=BRIEF)

## Levene's Test for Homogeneity of Variance (center = median)
##        Df F value  Pr(>F)    
## group   3  29.968 1.9e-14 ***
##       117                    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

AoLE_LangGrp <- kruskal.test(Age.of.Language.Exposure..mo.~LanguageGroup, data=BRIEF)
AoLE_LangGrp

## 
##  Kruskal-Wallis rank sum test
## 
## data:  Age.of.Language.Exposure..mo. by LanguageGroup
## Kruskal-Wallis chi-squared = 113.4, df = 3, p-value < 2.2e-16

dunnTest(Age.of.Language.Exposure..mo.~LanguageGroup, data=BRIEF, method="hochberg")

## Warning: Some rows deleted from 'x' and 'g' because missing data.

## Dunn (1964) Kruskal-Wallis multiple comparison

##   p-values adjusted with the Hochberg method.

##                          Comparison         Z      P.unadj        P.adj
## 1             Early ASL - Later ASL -7.894319 2.919049e-15 1.459524e-14
## 2         Early ASL - Later English -5.810672 6.222254e-09 1.866676e-08
## 3         Later ASL - Later English  1.838167 6.603775e-02 1.320755e-01
## 4     Early ASL - Typically Hearing  0.000000 1.000000e+00 1.000000e+00
## 5     Later ASL - Typically Hearing  8.923646 4.511844e-19 2.707106e-18
## 6 Later English - Typically Hearing  6.513165 7.358344e-11 2.943338e-10

Findings:

Both Early groups significantly different from both Later groups (as expected)
Later ASL group does not differ in age of language exposure compared to Later English group

Checking correlations between continuous (numerical) predictors

## Function below from: http://www.sthda.com/english/wiki/correlation-matrix-formatting-and-visualization
num_predictors <- BRIEF_AgeOf[, c("SES..3.66.", "AgeMonths", "Age.of.Language.Exposure..mo.", "AoAE")]
flattenCorrMatrix <- function(cormat, pmat) {
  ut <- upper.tri(cormat)
  data.frame(
    row = rownames(cormat)[row(cormat)[ut]],
    column = rownames(cormat)[col(cormat)[ut]],
    cor  =(cormat)[ut],
    p = pmat[ut]
    )
}
BRIEF_num_pred <- rcorr(as.matrix(num_predictors), type = c("pearson"))
flattenCorrMatrix(BRIEF_num_pred$r, BRIEF_num_pred$P)

##                             row                        column          cor
## 1                    SES..3.66.                     AgeMonths -0.002497801
## 2                    SES..3.66. Age.of.Language.Exposure..mo. -0.244521622
## 3                     AgeMonths Age.of.Language.Exposure..mo.  0.422545190
## 4                    SES..3.66.                          AoAE -0.211797298
## 5                     AgeMonths                          AoAE  0.399277901
## 6 Age.of.Language.Exposure..mo.                          AoAE  0.429319193
##              p
## 1 9.794351e-01
## 2 1.039299e-02
## 3 4.719230e-06
## 4 2.704404e-02
## 5 1.701768e-05
## 6 3.189754e-06

Findings:

SES is not significantly correlated with any other continuous predictors
Age correlated with both Age of Language Exposure and Age of Auditory Exposure (unsurprising)
Age of Language exposure significantly correlated with Age of Auditory Exposure (also unsurprising given that this will be correlated for participants learning spoken English, which is >50% of our sample)

Global Executive Composite Models

Step 1. Base Model (Demographic variables only):

overall_all_base <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths)
summary(overall_all_base)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -32.330 -14.855  -0.360   8.628  60.531 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 107.1580    11.5903   9.245 3.02e-15 ***
## SES..3.66.   -0.4345     0.1362  -3.191  0.00187 ** 
## SexMale      -1.1793     3.6935  -0.319  0.75014    
## AgeMonths     0.1167     0.1420   0.822  0.41320    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.94 on 105 degrees of freedom
## Multiple R-squared:  0.09454,    Adjusted R-squared:  0.06867 
## F-statistic: 3.654 on 3 and 105 DF,  p-value: 0.01492

AIC(overall_all_base)

## [1] 956.5089

Step 2. Add Age of Language Exposure:

overall_all_AoLE <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.)
summary(overall_all_AoLE)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo., 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -34.445 -13.455  -0.562   8.300  66.558 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   109.7397    11.2424   9.761  2.3e-16 ***
## SES..3.66.                     -0.3318     0.1364  -2.433  0.01667 *  
## SexMale                        -1.8352     3.5785  -0.513  0.60916    
## AgeMonths                      -0.0767     0.1529  -0.502  0.61687    
## Age.of.Language.Exposure..mo.   0.3104     0.1077   2.883  0.00479 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.32 on 104 degrees of freedom
## Multiple R-squared:  0.1615, Adjusted R-squared:  0.1293 
## F-statistic: 5.009 on 4 and 104 DF,  p-value: 0.0009866

AIC(overall_all_AoLE)

## [1] 950.1289

Compare Model from Step 2 to model from Step 1:

anova(overall_all_base, overall_all_AoLE)

## Analysis of Variance Table
## 
## Model 1: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
##   Res.Df   RSS Df Sum of Sq      F   Pr(>F)   
## 1    105 37686                                
## 2    104 34897  1    2788.7 8.3109 0.004789 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

3a. Add Age of Auditory Exposure:

overall_all_AoLE_AoAE <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+AoAE)
summary(overall_all_AoLE_AoAE)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -35.488 -13.012  -0.955   9.269  64.745 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   109.30858   11.24579   9.720 3.11e-16 ***
## SES..3.66.                     -0.35417    0.13801  -2.566  0.01172 *  
## SexMale                        -1.70684    3.57928  -0.477  0.63447    
## AgeMonths                      -0.02933    0.15945  -0.184  0.85440    
## Age.of.Language.Exposure..mo.   0.34130    0.11166   3.057  0.00285 ** 
## AoAE                           -0.08470    0.08152  -1.039  0.30121    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.31 on 103 degrees of freedom
## Multiple R-squared:  0.1702, Adjusted R-squared:   0.13 
## F-statistic: 4.226 on 5 and 103 DF,  p-value: 0.001553

AIC(overall_all_AoLE_AoAE)

## [1] 950.9923

3b. Add Length of Auditory Experience:

overall_all_AoLE_LoAE <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Length.of.Auditory.Exposure)
summary(overall_all_AoLE_LoAE)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Length.of.Auditory.Exposure, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -35.475 -13.038  -0.980   9.309  64.787 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   109.30137   11.24709   9.718 3.13e-16 ***
## SES..3.66.                     -0.35384    0.13800  -2.564  0.01179 *  
## SexMale                        -1.70954    3.57952  -0.478  0.63395    
## AgeMonths                      -0.11363    0.15695  -0.724  0.47071    
## Age.of.Language.Exposure..mo.   0.34085    0.11162   3.054  0.00288 ** 
## Length.of.Auditory.Exposure     0.08405    0.08153   1.031  0.30502    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.31 on 103 degrees of freedom
## Multiple R-squared:  0.1701, Adjusted R-squared:  0.1298 
## F-statistic: 4.222 on 5 and 103 DF,  p-value: 0.001565

AIC(overall_all_AoLE_LoAE)

## [1] 951.0101

3c. Add Modality:

overall_all_AoLE_Modality <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Language_Modality)
summary(overall_all_AoLE_Modality)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Language_Modality, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -34.164 -12.261  -1.154   8.590  65.839 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   110.23865   11.32017   9.738 2.83e-16 ***
## SES..3.66.                     -0.34263    0.13834  -2.477  0.01489 *  
## SexMale                        -1.82344    3.59098  -0.508  0.61269    
## AgeMonths                      -0.06616    0.15465  -0.428  0.66967    
## Age.of.Language.Exposure..mo.   0.32718    0.11254   2.907  0.00447 ** 
## Language_ModalityASL           -2.15194    4.04013  -0.533  0.59543    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.38 on 103 degrees of freedom
## Multiple R-squared:  0.1638, Adjusted R-squared:  0.1233 
## F-statistic: 4.037 on 5 and 103 DF,  p-value: 0.002191

AIC(overall_all_AoLE_Modality)

## [1] 951.8291

3d.Add Hearing Status (3-category):

overall_all_AoLE_HearStat <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+ Hearing_Level_3_Cat)
summary(overall_all_AoLE_HearStat)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Hearing_Level_3_Cat, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -33.895 -12.842  -0.688   8.396  65.922 
## 
## Coefficients:
##                                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                        112.3796    12.2077   9.206 7.57e-15 ***
## SES..3.66.                          -0.3348     0.1430  -2.341   0.0213 *  
## SexMale                             -0.6712     3.7204  -0.180   0.8572    
## AgeMonths                           -0.1033     0.1686  -0.613   0.5415    
## Age.of.Language.Exposure..mo.        0.3130     0.1316   2.378   0.0194 *  
## Hearing_Level_3_CatDeaf             -2.3984     4.9761  -0.482   0.6309    
## Hearing_Level_3_CatHard of Hearing  -2.2487     6.3660  -0.353   0.7247    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.48 on 96 degrees of freedom
##   (6 observations deleted due to missingness)
## Multiple R-squared:  0.1434, Adjusted R-squared:  0.08983 
## F-statistic: 2.678 on 6 and 96 DF,  p-value: 0.01905

AIC(overall_all_AoLE_HearStat)

## [1] 901.8779

Compare Models from Step 3a-c to model from Step 2:

anova(overall_all_AoLE, overall_all_AoLE_AoAE)

## Analysis of Variance Table
## 
## Model 1: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     AoAE
##   Res.Df   RSS Df Sum of Sq      F Pr(>F)
## 1    104 34897                           
## 2    103 34535  1    361.99 1.0796 0.3012

anova(overall_all_AoLE, overall_all_AoLE_Modality)

## Analysis of Variance Table
## 
## Model 1: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Language_Modality
##   Res.Df   RSS Df Sum of Sq      F Pr(>F)
## 1    104 34897                           
## 2    103 34801  1    95.857 0.2837 0.5954

anova(overall_all_AoLE, overall_all_AoLE_LoAE)

## Analysis of Variance Table
## 
## Model 1: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Length.of.Auditory.Exposure
##   Res.Df   RSS Df Sum of Sq      F Pr(>F)
## 1    104 34897                           
## 2    103 34541  1    356.36 1.0627  0.305

Findings: Age of auditory exposure, length of auditory exposure, and language modality are not significant predictors of GEC raw scores, and their addition does not improve model fit.

We cannot do a direct comparison of model in step 3d to the model in step 2 because we are missing hearing status information for 12 participants (mostly Early and Later ASL). However, even in the model in step 3d, Hearing Status is not a significant predictor of BRIEF scores, and Age of Language Exposure is.

Checking assumptions for best-fit GEC model

ols_plot_resid_qq(overall_all_AoLE)

ols_test_normality(overall_all_AoLE)

## -----------------------------------------------
##        Test             Statistic       pvalue  
## -----------------------------------------------
## Shapiro-Wilk              0.9674         0.0089 
## Kolmogorov-Smirnov        0.0841         0.4242 
## Cramer-von Mises          8.634          0.0000 
## Anderson-Darling          0.8395         0.0296 
## -----------------------------------------------

ols_plot_resid_hist(overall_all_AoLE)

ols_test_correlation(overall_all_AoLE)

## [1] 0.9826969

ols_plot_resid_fit(overall_all_AoLE)

Inhibition Models

Step 1. Base Model:

inhibition_all_base <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths)
summary(inhibition_all_base)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.4961  -4.5032   0.0108   4.0216  21.5107 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 25.07731    3.65405   6.863 4.86e-10 ***
## SES..3.66.  -0.10990    0.04292  -2.560   0.0119 *  
## SexMale      0.97828    1.16445   0.840   0.4028    
## AgeMonths    0.06696    0.04478   1.495   0.1378    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.973 on 105 degrees of freedom
## Multiple R-squared:  0.08539,    Adjusted R-squared:  0.05926 
## F-statistic: 3.268 on 3 and 105 DF,  p-value: 0.02424

AIC(inhibition_all_base)

## [1] 704.8643

Step 2. Add Age of Language Exposure:

inhibition_all_AoLE <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.)
summary(inhibition_all_AoLE)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.9261  -4.2881  -0.1993   4.2050  22.7360 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.60218    3.62617   7.060 1.93e-10 ***
## SES..3.66.                    -0.08903    0.04399  -2.024   0.0455 *  
## SexMale                        0.84495    1.15423   0.732   0.4658    
## AgeMonths                      0.02764    0.04930   0.561   0.5762    
## Age.of.Language.Exposure..mo.  0.06310    0.03473   1.817   0.0721 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.908 on 104 degrees of freedom
## Multiple R-squared:  0.1135, Adjusted R-squared:  0.07944 
## F-statistic:  3.33 on 4 and 104 DF,  p-value: 0.01311

AIC(inhibition_all_AoLE)

## [1] 703.4576

Compare Model with Age of Language Exposure to base demographic model:

anova(inhibition_all_base, inhibition_all_AoLE)

## Analysis of Variance Table
## 
## Model 1: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F  Pr(>F)  
## 1    105 3745.7                              
## 2    104 3630.5  1    115.26 3.3018 0.07208 .
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

3a. Add Age of Auditory Experience:

inhibition_all_AoLE_AoAE <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+AoAE)
summary(inhibition_all_AoLE_AoAE)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.3417  -4.3176  -0.1074   4.1316  22.2424 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.48478    3.63271   7.015 2.49e-10 ***
## SES..3.66.                    -0.09511    0.04458  -2.133   0.0353 *  
## SexMale                        0.87988    1.15621   0.761   0.4484    
## AgeMonths                      0.04054    0.05151   0.787   0.4330    
## Age.of.Language.Exposure..mo.  0.07152    0.03607   1.983   0.0501 .  
## AoAE                          -0.02306    0.02633  -0.876   0.3832    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.915 on 103 degrees of freedom
## Multiple R-squared:  0.1201, Adjusted R-squared:  0.07737 
## F-statistic: 2.811 on 5 and 103 DF,  p-value: 0.02018

AIC(inhibition_all_AoLE_AoAE)

## [1] 704.6489

3b. Add Length of Auditory Exposure:

inhibition_all_AoLE_LoAE <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Length.of.Auditory.Exposure)
summary(inhibition_all_AoLE_LoAE)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Length.of.Auditory.Exposure, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.3466  -4.3172  -0.0975   4.1237  22.2565 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.48347    3.63318   7.014  2.5e-10 ***
## SES..3.66.                    -0.09498    0.04458  -2.131   0.0355 *  
## SexMale                        0.87896    1.15630   0.760   0.4489    
## AgeMonths                      0.01764    0.05070   0.348   0.7285    
## Age.of.Language.Exposure..mo.  0.07135    0.03606   1.979   0.0505 .  
## Length.of.Auditory.Exposure    0.02276    0.02634   0.864   0.3895    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.916 on 103 degrees of freedom
## Multiple R-squared:  0.1199, Adjusted R-squared:  0.07719 
## F-statistic: 2.807 on 5 and 103 DF,  p-value: 0.02035

AIC(inhibition_all_AoLE_LoAE)

## [1] 704.6701

3c. Add Modality:

inhibition_all_AoLE_Modality <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+ Language_Modality)
summary(inhibition_all_AoLE_Modality)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Language_Modality, data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.9319  -4.2811  -0.2139   4.2162  22.7509 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.59185    3.65624   6.999 2.69e-10 ***
## SES..3.66.                    -0.08880    0.04468  -1.987   0.0495 *  
## SexMale                        0.84470    1.15983   0.728   0.4681    
## AgeMonths                      0.02743    0.04995   0.549   0.5841    
## Age.of.Language.Exposure..mo.  0.06276    0.03635   1.726   0.0873 .  
## Language_ModalityASL           0.04455    1.30490   0.034   0.9728    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.937 on 103 degrees of freedom
## Multiple R-squared:  0.1135, Adjusted R-squared:  0.07051 
## F-statistic: 2.639 on 5 and 103 DF,  p-value: 0.02752

AIC(inhibition_all_AoLE_Modality)

## [1] 705.4563

3d. Add Hearing Status (3-category):

inhibition_all_AoLE_HearStat <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+ Hearing_Level_3_Cat)
summary(inhibition_all_AoLE_HearStat)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Hearing_Level_3_Cat, data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.6102  -4.5798  -0.0519   3.7536  22.9804 
## 
## Coefficients:
##                                     Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                        26.361868   3.902861   6.754  1.1e-09 ***
## SES..3.66.                         -0.082412   0.045729  -1.802   0.0747 .  
## SexMale                             1.051729   1.189446   0.884   0.3788    
## AgeMonths                           0.008764   0.053899   0.163   0.8712    
## Age.of.Language.Exposure..mo.       0.049904   0.042073   1.186   0.2385    
## Hearing_Level_3_CatDeaf             0.620825   1.590893   0.390   0.6972    
## Hearing_Level_3_CatHard of Hearing -1.545955   2.035248  -0.760   0.4494    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.908 on 96 degrees of freedom
##   (6 observations deleted due to missingness)
## Multiple R-squared:  0.1096, Adjusted R-squared:  0.0539 
## F-statistic: 1.969 on 6 and 96 DF,  p-value: 0.07767

AIC(inhibition_all_AoLE_HearStat)

## [1] 666.9651

Compare Models from Step 3a-c to model from Step 2:

anova(inhibition_all_AoLE, inhibition_all_AoLE_AoAE)

## Analysis of Variance Table
## 
## Model 1: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 3630.5                           
## 2    103 3603.6  1    26.837 0.7671 0.3832

anova(inhibition_all_AoLE, inhibition_all_AoLE_Modality)

## Analysis of Variance Table
## 
## Model 1: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Language_Modality
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 3630.5                           
## 2    103 3630.4  1  0.041089 0.0012 0.9728

anova(inhibition_all_AoLE, inhibition_all_AoLE_LoAE)

## Analysis of Variance Table
## 
## Model 1: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Length.of.Auditory.Exposure
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 3630.5                           
## 2    103 3604.3  1    26.134 0.7468 0.3895

Checking assumptions for best-fit Inhibition model

ols_plot_resid_qq(inhibition_all_AoLE)

ols_test_normality(inhibition_all_AoLE)

## -----------------------------------------------
##        Test             Statistic       pvalue  
## -----------------------------------------------
## Shapiro-Wilk              0.9723         0.0226 
## Kolmogorov-Smirnov        0.0732         0.6029 
## Cramer-von Mises          8.308          0.0000 
## Anderson-Darling          0.5534         0.1502 
## -----------------------------------------------

ols_plot_resid_hist(inhibition_all_AoLE)

ols_test_correlation(inhibition_all_AoLE)

## [1] 0.9845734

ols_plot_resid_fit(inhibition_all_AoLE)

Shift Models

Step 1. Base Model:

shift_all_base <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths)
summary(shift_all_base)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths, data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -4.936 -2.682 -0.776  1.774 14.985 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 17.93727    2.20373   8.140 8.65e-13 ***
## SES..3.66.  -0.08380    0.02589  -3.237  0.00162 ** 
## SexMale     -0.79353    0.70227  -1.130  0.26107    
## AgeMonths    0.01077    0.02701   0.399  0.69088    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.602 on 105 degrees of freedom
## Multiple R-squared:  0.0979, Adjusted R-squared:  0.07213 
## F-statistic: 3.799 on 3 and 105 DF,  p-value: 0.01245

AIC(shift_all_base)

## [1] 594.6247

Step 2. Add Age of Language Exposure:

shift_all_AoLE <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.)
summary(shift_all_AoLE)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.5170 -2.2586 -0.6156  1.4734 14.0409 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   18.43971    2.13351   8.643 7.11e-14 ***
## SES..3.66.                    -0.06382    0.02588  -2.466  0.01530 *  
## SexMale                       -0.92116    0.67911  -1.356  0.17790    
## AgeMonths                     -0.02687    0.02901  -0.926  0.35646    
## Age.of.Language.Exposure..mo.  0.06041    0.02043   2.956  0.00385 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.476 on 104 degrees of freedom
## Multiple R-squared:  0.1678, Adjusted R-squared:  0.1358 
## F-statistic: 5.244 on 4 and 104 DF,  p-value: 0.0006899

AIC(shift_all_AoLE)

## [1] 587.829

Compare Model with Age of Language Exposure to base demographic model:

anova(shift_all_base, shift_all_AoLE)

## Analysis of Variance Table
## 
## Model 1: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F   Pr(>F)   
## 1    105 1362.4                                
## 2    104 1256.8  1    105.62 8.7401 0.003852 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

3a. Add Age of Auditory Experience:

shift_all_AoLE_AoAE <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+AoAE)
summary(shift_all_AoLE_AoAE)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.4369 -2.2772 -0.5143  1.4576 14.0693 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   18.425943   2.144990   8.590 9.89e-14 ***
## SES..3.66.                    -0.064533   0.026323  -2.452   0.0159 *  
## SexMale                       -0.917061   0.682702  -1.343   0.1821    
## AgeMonths                     -0.025355   0.030413  -0.834   0.4064    
## Age.of.Language.Exposure..mo.  0.061394   0.021298   2.883   0.0048 ** 
## AoAE                          -0.002704   0.015549  -0.174   0.8623    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.493 on 103 degrees of freedom
## Multiple R-squared:  0.1681, Adjusted R-squared:  0.1277 
## F-statistic: 4.162 on 5 and 103 DF,  p-value: 0.001745

AIC(shift_all_AoLE_AoAE)

## [1] 589.797

3b. Add Length of Auditory Exposure:

shift_all_AoLE_LoAE <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Length.of.Auditory.Exposure)  
summary(shift_all_AoLE_LoAE)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Length.of.Auditory.Exposure, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.4408 -2.2740 -0.5135  1.4587 14.0684 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   18.426317   2.145094   8.590 9.91e-14 ***
## SES..3.66.                    -0.064492   0.026319  -2.450  0.01596 *  
## SexMale                       -0.917321   0.682701  -1.344  0.18201    
## AgeMonths                     -0.027996   0.029934  -0.935  0.35185    
## Age.of.Language.Exposure..mo.  0.061337   0.021288   2.881  0.00482 ** 
## Length.of.Auditory.Exposure    0.002567   0.015550   0.165  0.86918    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.493 on 103 degrees of freedom
## Multiple R-squared:  0.1681, Adjusted R-squared:  0.1277 
## F-statistic: 4.161 on 5 and 103 DF,  p-value: 0.001748

AIC(shift_all_AoLE_LoAE)

## [1] 589.8002

3c. Add Modality:

shift_all_AoLE_Modality <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Language_Modality)  
summary(shift_all_AoLE_Modality)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Language_Modality, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.5425 -2.2325 -0.6918  1.4192 14.0009 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   18.41341    2.15099   8.560 1.15e-13 ***
## SES..3.66.                    -0.06325    0.02629  -2.406   0.0179 *  
## SexMale                       -0.92178    0.68234  -1.351   0.1797    
## AgeMonths                     -0.02742    0.02939  -0.933   0.3529    
## Age.of.Language.Exposure..mo.  0.05952    0.02138   2.783   0.0064 ** 
## Language_ModalityASL           0.11343    0.76768   0.148   0.8828    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.493 on 103 degrees of freedom
## Multiple R-squared:  0.168,  Adjusted R-squared:  0.1276 
## F-statistic:  4.16 on 5 and 103 DF,  p-value: 0.001752

AIC(shift_all_AoLE_Modality)

## [1] 589.8059

3d. Add Hearing Status (3-category):

shift_all_AoLE_HearStat <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+ Hearing_Level_3_Cat)
summary(shift_all_AoLE_HearStat)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Hearing_Level_3_Cat, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.9789 -2.2795 -0.4574  1.7698 12.7067 
## 
## Coefficients:
##                                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                        18.39764    2.30188   7.992 2.95e-12 ***
## SES..3.66.                         -0.07108    0.02697  -2.635   0.0098 ** 
## SexMale                            -0.69119    0.70153  -0.985   0.3270    
## AgeMonths                          -0.01995    0.03179  -0.628   0.5318    
## Age.of.Language.Exposure..mo.       0.04517    0.02481   1.820   0.0718 .  
## Hearing_Level_3_CatDeaf            -0.19032    0.93830  -0.203   0.8397    
## Hearing_Level_3_CatHard of Hearing  1.76805    1.20037   1.473   0.1440    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.484 on 96 degrees of freedom
##   (6 observations deleted due to missingness)
## Multiple R-squared:  0.1804, Adjusted R-squared:  0.1291 
## F-statistic: 3.521 on 6 and 96 DF,  p-value: 0.003423

AIC(shift_all_AoLE_HearStat)

## [1] 558.2004

Compare Models from Step 3a-c to model from Step 2:

anova(shift_all_AoLE, shift_all_AoLE_AoAE)

## Analysis of Variance Table
## 
## Model 1: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1256.8                           
## 2    103 1256.4  1     0.369 0.0303 0.8623

anova(shift_all_AoLE, shift_all_AoLE_Modality)

## Analysis of Variance Table
## 
## Model 1: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Language_Modality
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1256.8                           
## 2    103 1256.5  1   0.26633 0.0218 0.8828

anova(shift_all_AoLE, shift_all_AoLE_LoAE)

## Analysis of Variance Table
## 
## Model 1: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Length.of.Auditory.Exposure
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1256.8                           
## 2    103 1256.4  1   0.33256 0.0273 0.8692

Checking assumptions for best-fit Shift model

ols_plot_resid_qq(shift_all_AoLE)

ols_test_normality(shift_all_AoLE)

## -----------------------------------------------
##        Test             Statistic       pvalue  
## -----------------------------------------------
## Shapiro-Wilk              0.9237         0.0000 
## Kolmogorov-Smirnov        0.0984         0.2423 
## Cramer-von Mises          9.0788         0.0000 
## Anderson-Darling          1.7709          1e-04 
## -----------------------------------------------

ols_plot_resid_hist(shift_all_AoLE)

ols_test_correlation(shift_all_AoLE)

## [1] 0.9589466

ols_plot_resid_fit(shift_all_AoLE)

Emotional Control Models

Step 1. Base Model:

emotctrl_all_base <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths)
summary(emotctrl_all_base)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -8.4561 -3.2273 -0.6821  2.4962 13.8386 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 20.049467   2.498494   8.025 1.55e-12 ***
## SES..3.66.  -0.091365   0.029350  -3.113  0.00239 ** 
## SexMale     -0.705779   0.796204  -0.886  0.37741    
## AgeMonths    0.001859   0.030620   0.061  0.95170    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.084 on 105 degrees of freedom
## Multiple R-squared:  0.08711,    Adjusted R-squared:  0.06103 
## F-statistic:  3.34 on 3 and 105 DF,  p-value: 0.02214

Step 2. Add Age of Language Exposure:

emotctrl_all_AoLE <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.)
summary(emotctrl_all_AoLE)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths + Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -8.1523 -3.0899 -0.5234  2.6179 13.4122 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   20.27630    2.50295   8.101 1.11e-12 ***
## SES..3.66.                    -0.08234    0.03036  -2.712  0.00783 ** 
## SexMale                       -0.76340    0.79670  -0.958  0.34018    
## AgeMonths                     -0.01513    0.03403  -0.445  0.65747    
## Age.of.Language.Exposure..mo.  0.02727    0.02397   1.138  0.25786    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.078 on 104 degrees of freedom
## Multiple R-squared:  0.09833,    Adjusted R-squared:  0.06365 
## F-statistic: 2.836 on 4 and 104 DF,  p-value: 0.0281

Compare Model with Age of Language Exposure to base demographic model:

anova(emotctrl_all_base, emotctrl_all_AoLE)

## Analysis of Variance Table
## 
## Model 1: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    105 1751.2                           
## 2    104 1729.7  1    21.528 1.2944 0.2579

3a. Add Age of Auditory Experience:

emotctrl_all_AoLE_AoAE <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+AoAE)
summary(emotctrl_all_AoLE_AoAE)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths + Age.of.Language.Exposure..mo. + AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -8.418 -3.282 -0.533  2.551 13.559 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   20.204803   2.509534   8.051  1.5e-12 ***
## SES..3.66.                    -0.086045   0.030797  -2.794  0.00621 ** 
## SexMale                       -0.742122   0.798728  -0.929  0.35499    
## AgeMonths                     -0.007278   0.035581  -0.205  0.83833    
## Age.of.Language.Exposure..mo.  0.032397   0.024918   1.300  0.19645    
## AoAE                          -0.014046   0.018191  -0.772  0.44179    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.086 on 103 degrees of freedom
## Multiple R-squared:  0.1035, Adjusted R-squared:   0.06 
## F-statistic: 2.379 on 5 and 103 DF,  p-value: 0.0437

3b. Add Length of Auditory Experience:

emotctrl_all_AoLE_LoAE <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Length.of.Auditory.Exposure)
summary(emotctrl_all_AoLE_LoAE)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths + Age.of.Language.Exposure..mo. + Length.of.Auditory.Exposure, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -8.421 -3.278 -0.527  2.557 13.562 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   20.20351    2.50971   8.050 1.51e-12 ***
## SES..3.66.                    -0.08599    0.03079  -2.793  0.00623 ** 
## SexMale                       -0.74254    0.79875  -0.930  0.35473    
## AgeMonths                     -0.02127    0.03502  -0.607  0.54506    
## Age.of.Language.Exposure..mo.  0.03233    0.02491   1.298  0.19719    
## Length.of.Auditory.Exposure    0.01396    0.01819   0.767  0.44474    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.086 on 103 degrees of freedom
## Multiple R-squared:  0.1035, Adjusted R-squared:  0.05994 
## F-statistic: 2.377 on 5 and 103 DF,  p-value: 0.04383

3c. Add Modality:

emotctrl_all_AoLE_Modality <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Language_Modality)
summary(emotctrl_all_AoLE_Modality)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths + Age.of.Language.Exposure..mo. + Language_Modality, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -8.2712 -3.1432 -0.4244  2.5967 13.4687 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   20.31350    2.52334   8.050 1.51e-12 ***
## SES..3.66.                    -0.08315    0.03084  -2.696  0.00819 ** 
## SexMale                       -0.76253    0.80045  -0.953  0.34301    
## AgeMonths                     -0.01435    0.03447  -0.416  0.67813    
## Age.of.Language.Exposure..mo.  0.02852    0.02509   1.137  0.25818    
## Language_ModalityASL          -0.16042    0.90057  -0.178  0.85897    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.097 on 103 degrees of freedom
## Multiple R-squared:  0.09861,    Adjusted R-squared:  0.05485 
## F-statistic: 2.254 on 5 and 103 DF,  p-value: 0.0545

3d. Add Hearing Status (3-category):

emotctrl_all_AoLE_HearStat <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+ Hearing_Level_3_Cat)
summary(emotctrl_all_AoLE_HearStat)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths + Age.of.Language.Exposure..mo. + Hearing_Level_3_Cat, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -7.7400 -2.9737 -0.7746  2.3407 13.1880 
## 
## Coefficients:
##                                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                        20.89805    2.71573   7.695 1.25e-11 ***
## SES..3.66.                         -0.08644    0.03182  -2.716  0.00783 ** 
## SexMale                            -0.56179    0.82765  -0.679  0.49892    
## AgeMonths                          -0.01541    0.03750  -0.411  0.68210    
## Age.of.Language.Exposure..mo.       0.03813    0.02928   1.302  0.19593    
## Hearing_Level_3_CatDeaf            -1.24135    1.10699  -1.121  0.26493    
## Hearing_Level_3_CatHard of Hearing -0.59943    1.41619  -0.423  0.67304    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.111 on 96 degrees of freedom
##   (6 observations deleted due to missingness)
## Multiple R-squared:  0.09777,    Adjusted R-squared:  0.04138 
## F-statistic: 1.734 on 6 and 96 DF,  p-value: 0.1213

Compare Models from Step 3a-c to model from Step 2:

anova(emotctrl_all_AoLE, emotctrl_all_AoLE_AoAE)

## Analysis of Variance Table
## 
## Model 1: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1729.7                           
## 2    103 1719.8  1    9.9548 0.5962 0.4418

anova(emotctrl_all_AoLE, emotctrl_all_AoLE_Modality)

## Analysis of Variance Table
## 
## Model 1: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Language_Modality
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1729.7                           
## 2    103 1729.2  1   0.53271 0.0317  0.859

anova(emotctrl_all_AoLE, emotctrl_all_AoLE_LoAE)

## Analysis of Variance Table
## 
## Model 1: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Length.of.Auditory.Exposure
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1729.7                           
## 2    103 1719.9  1    9.8272 0.5885 0.4447

Checking assumptions for best-fit Emotional Control model

ols_plot_resid_qq(emotctrl_all_AoLE)

ols_test_normality(emotctrl_all_AoLE)

## -----------------------------------------------
##        Test             Statistic       pvalue  
## -----------------------------------------------
## Shapiro-Wilk              0.9536          8e-04 
## Kolmogorov-Smirnov        0.1098         0.1447 
## Cramer-von Mises         10.0351         0.0000 
## Anderson-Darling          1.4899          7e-04 
## -----------------------------------------------

ols_plot_resid_hist(emotctrl_all_AoLE)

ols_test_correlation(emotctrl_all_AoLE)

## [1] 0.9762239

ols_plot_resid_fit(emotctrl_all_AoLE)

Working Memory Models

Step 1. Base Model:

wm_all_base <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths)
summary(wm_all_base)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -8.748 -4.607 -1.414  3.432 23.472 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 25.20501    3.70291   6.807 6.38e-10 ***
## SES..3.66.  -0.07306    0.04350  -1.680    0.096 .  
## SexMale     -0.43333    1.18002  -0.367    0.714    
## AgeMonths    0.04043    0.04538   0.891    0.375    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 6.053 on 105 degrees of freedom
## Multiple R-squared:  0.0346, Adjusted R-squared:  0.007015 
## F-statistic: 1.254 on 3 and 105 DF,  p-value: 0.294

AIC(wm_all_base)

## [1] 707.7603

Step 2. Add Age of Language Exposure:

wm_all_AoLE <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.)
summary(wm_all_AoLE)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -9.445 -4.061 -1.545  3.381 25.457 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   26.05515    3.58282   7.272 6.86e-11 ***
## SES..3.66.                    -0.03925    0.04346  -0.903   0.3686    
## SexMale                       -0.64928    1.14043  -0.569   0.5704    
## AgeMonths                     -0.02326    0.04871  -0.477   0.6341    
## Age.of.Language.Exposure..mo.  0.10221    0.03431   2.979   0.0036 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.838 on 104 degrees of freedom
## Multiple R-squared:  0.1105, Adjusted R-squared:  0.07628 
## F-statistic:  3.23 on 4 and 104 DF,  p-value: 0.01531

AIC(wm_all_AoLE)

## [1] 700.8361

Compare Model with Age of Language Exposure to base demographic model:

anova(wm_all_base, wm_all_AoLE)

## Analysis of Variance Table
## 
## Model 1: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F   Pr(>F)   
## 1    105 3846.6                                
## 2    104 3544.2  1    302.39 8.8732 0.003602 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

3a. Add Age of Auditory Experience:

wm_all_AoLE_AoAE <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+AoAE)
summary(wm_all_AoLE_AoAE)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -9.313 -4.115 -1.080  3.208 24.929 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.929624   3.586996   7.229 8.82e-11 ***
## SES..3.66.                    -0.045745   0.044020  -1.039  0.30115    
## SexMale                       -0.611926   1.141660  -0.536  0.59312    
## AgeMonths                     -0.009465   0.050858  -0.186  0.85272    
## Age.of.Language.Exposure..mo.  0.111208   0.035616   3.122  0.00233 ** 
## AoAE                          -0.024660   0.026002  -0.948  0.34514    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.841 on 103 degrees of freedom
## Multiple R-squared:  0.1182, Adjusted R-squared:  0.07538 
## F-statistic: 2.761 on 5 and 103 DF,  p-value: 0.02209

AIC(wm_all_AoLE_AoAE)

## [1] 701.8883

3b. Add Length of Auditory Experience:

wm_all_AoLE_LoAE <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Length.of.Auditory.Exposure)
summary(wm_all_AoLE_LoAE)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Length.of.Auditory.Exposure, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -9.311 -4.115 -1.085  3.214 24.940 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.92725    3.58730   7.228 8.87e-11 ***
## SES..3.66.                    -0.04566    0.04401  -1.037  0.30197    
## SexMale                       -0.61263    1.14170  -0.537  0.59270    
## AgeMonths                     -0.03403    0.05006  -0.680  0.49816    
## Age.of.Language.Exposure..mo.  0.11109    0.03560   3.121  0.00234 ** 
## Length.of.Auditory.Exposure    0.02452    0.02600   0.943  0.34791    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.841 on 103 degrees of freedom
## Multiple R-squared:  0.1181, Adjusted R-squared:  0.07529 
## F-statistic: 2.759 on 5 and 103 DF,  p-value: 0.02219

AIC(wm_all_AoLE_LoAE)

## [1] 701.8992

3c. Add Modality:

wm_all_AoLE_Modality <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Language_Modality)
summary(wm_all_AoLE_Modality)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Language_Modality, data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -9.328 -4.082 -1.427  3.255 25.158 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   26.26220    3.60414   7.287 6.64e-11 ***
## SES..3.66.                    -0.04372    0.04404  -0.993  0.32322    
## SexMale                       -0.64442    1.14330  -0.564  0.57422    
## AgeMonths                     -0.01888    0.04924  -0.384  0.70213    
## Age.of.Language.Exposure..mo.  0.10918    0.03583   3.047  0.00294 ** 
## Language_ModalityASL          -0.89303    1.28630  -0.694  0.48908    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.852 on 103 degrees of freedom
## Multiple R-squared:  0.1146, Adjusted R-squared:  0.07165 
## F-statistic: 2.667 on 5 and 103 DF,  p-value: 0.02615

AIC(wm_all_AoLE_Modality)

## [1] 702.3272

3d. Add Hearing Status (3-category):

wm_all_AoLE_HearStat <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+ Hearing_Level_3_Cat)
summary(wm_all_AoLE_HearStat)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Hearing_Level_3_Cat, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -9.4405 -4.3817 -0.8562  2.9612 25.3821 
## 
## Coefficients:
##                                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                        26.69333    3.89045   6.861 6.67e-10 ***
## SES..3.66.                         -0.03277    0.04558  -0.719  0.47391    
## SexMale                            -0.41046    1.18566  -0.346  0.72996    
## AgeMonths                          -0.03575    0.05373  -0.665  0.50742    
## Age.of.Language.Exposure..mo.       0.11179    0.04194   2.665  0.00902 ** 
## Hearing_Level_3_CatDeaf            -0.44075    1.58583  -0.278  0.78167    
## Hearing_Level_3_CatHard of Hearing -1.82749    2.02878  -0.901  0.36996    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.889 on 96 degrees of freedom
##   (6 observations deleted due to missingness)
## Multiple R-squared:  0.1026, Adjusted R-squared:  0.04653 
## F-statistic:  1.83 on 6 and 96 DF,  p-value: 0.1013

AIC(wm_all_AoLE_HearStat)

## [1] 666.309

Compare Models from Step 3a-c to model from Step 2:

anova(wm_all_AoLE, wm_all_AoLE_AoAE)

## Analysis of Variance Table
## 
## Model 1: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 3544.2                           
## 2    103 3513.5  1    30.683 0.8995 0.3451

anova(wm_all_AoLE, wm_all_AoLE_Modality)

## Analysis of Variance Table
## 
## Model 1: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Language_Modality
##   Res.Df    RSS Df Sum of Sq     F Pr(>F)
## 1    104 3544.2                          
## 2    103 3527.7  1    16.508 0.482 0.4891

anova(wm_all_AoLE, wm_all_AoLE_LoAE)

## Analysis of Variance Table
## 
## Model 1: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Length.of.Auditory.Exposure
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 3544.2                           
## 2    103 3513.9  1    30.334 0.8892 0.3479

Checking assumptions for best-fit Working Memory model

ols_plot_resid_qq(wm_all_AoLE)

ols_test_normality(wm_all_AoLE)

## -----------------------------------------------
##        Test             Statistic       pvalue  
## -----------------------------------------------
## Shapiro-Wilk              0.9173         0.0000 
## Kolmogorov-Smirnov        0.1184         0.0940 
## Cramer-von Mises          9.025          0.0000 
## Anderson-Darling          1.8261          1e-04 
## -----------------------------------------------

ols_plot_resid_hist(wm_all_AoLE)

ols_test_correlation(wm_all_AoLE)

## [1] 0.9554127

ols_plot_resid_fit(wm_all_AoLE)

Planning/Organization Models

Step 1. Base Model:

plan_all_base <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths)
summary(plan_all_base)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.2655 -2.8630 -0.4688  2.5755  8.6681 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 18.888939   2.203955   8.570 9.66e-14 ***
## SES..3.66.  -0.076399   0.025890  -2.951  0.00391 ** 
## SexMale     -0.224973   0.702342  -0.320  0.74936    
## AgeMonths   -0.003322   0.027010  -0.123  0.90234    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.602 on 105 degrees of freedom
## Multiple R-squared:  0.07663,    Adjusted R-squared:  0.05024 
## F-statistic: 2.904 on 3 and 105 DF,  p-value: 0.03826

AIC(plan_all_base)

## [1] 594.6475

Step 2. Add Age of Language Exposure:

plan_all_AoLE <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.)
summary(plan_all_AoLE)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.6566 -2.5795 -0.5434  2.0855  9.1255 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   19.36639    2.14243   9.039 9.37e-15 ***
## SES..3.66.                    -0.05741    0.02599  -2.209  0.02937 *  
## SexMale                       -0.34626    0.68194  -0.508  0.61270    
## AgeMonths                     -0.03909    0.02913  -1.342  0.18254    
## Age.of.Language.Exposure..mo.  0.05740    0.02052   2.798  0.00613 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.491 on 104 degrees of freedom
## Multiple R-squared:  0.1413, Adjusted R-squared:  0.1082 
## F-statistic: 4.277 on 4 and 104 DF,  p-value: 0.003037

AIC(plan_all_AoLE)

## [1] 588.7381

Compare Model with Age of Language Exposure to base demographic model:

anova(plan_all_base, plan_all_AoLE)

## Analysis of Variance Table
## 
## Model 1: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F   Pr(>F)   
## 1    105 1362.7                                
## 2    104 1267.3  1    95.378 7.8271 0.006134 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

3a. Add Age of Auditory Experience:

plan_all_AoLE_AoAE <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+AoAE)
summary(plan_all_AoLE_AoAE)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.3537 -2.3268 -0.6309  2.5441  8.6925 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   19.26342    2.13665   9.016 1.14e-14 ***
## SES..3.66.                    -0.06274    0.02622  -2.393  0.01853 *  
## SexMale                       -0.31561    0.68005  -0.464  0.64355    
## AgeMonths                     -0.02778    0.03029  -0.917  0.36136    
## Age.of.Language.Exposure..mo.  0.06478    0.02122   3.054  0.00288 ** 
## AoAE                          -0.02023    0.01549  -1.306  0.19443    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.479 on 103 degrees of freedom
## Multiple R-squared:  0.1552, Adjusted R-squared:  0.1142 
## F-statistic: 3.786 on 5 and 103 DF,  p-value: 0.003455

AIC(plan_all_AoLE_AoAE)

## [1] 588.9476

3b. Add Length of Auditory Experience:

plan_all_AoLE_LoAE <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Length.of.Auditory.Exposure)
summary(plan_all_AoLE_LoAE)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Length.of.Auditory.Exposure, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.3527 -2.3279 -0.6324  2.5437  8.6990 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   19.26082    2.13670   9.014 1.15e-14 ***
## SES..3.66.                    -0.06271    0.02622  -2.392  0.01858 *  
## SexMale                       -0.31601    0.68003  -0.465  0.64313    
## AgeMonths                     -0.04798    0.02982  -1.609  0.11063    
## Age.of.Language.Exposure..mo.  0.06474    0.02120   3.053  0.00288 ** 
## Length.of.Auditory.Exposure    0.02024    0.01549   1.307  0.19420    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.479 on 103 degrees of freedom
## Multiple R-squared:  0.1553, Adjusted R-squared:  0.1143 
## F-statistic: 3.786 on 5 and 103 DF,  p-value: 0.003452

AIC(plan_all_AoLE_LoAE)

## [1] 588.9458

3c. Add Modality:

plan_all_AoLE_Modality <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+Language_Modality)
summary(plan_all_AoLE_Modality)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Language_Modality, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.4928 -2.4645 -0.5671  2.2852  8.7057 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   19.65770    2.13218   9.220 4.02e-15 ***
## SES..3.66.                    -0.06371    0.02606  -2.445  0.01619 *  
## SexMale                       -0.33942    0.67637  -0.502  0.61686    
## AgeMonths                     -0.03294    0.02913  -1.131  0.26080    
## Age.of.Language.Exposure..mo.  0.06720    0.02120   3.170  0.00201 ** 
## Language_ModalityASL          -1.25648    0.76097  -1.651  0.10175    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.462 on 103 degrees of freedom
## Multiple R-squared:  0.1634, Adjusted R-squared:  0.1228 
## F-statistic: 4.023 on 5 and 103 DF,  p-value: 0.002244

AIC(plan_all_AoLE_Modality)

## [1] 587.8905

3d. Add Hearing Status (3-category):

plan_all_AoLE_HearStat <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths+Age.of.Language.Exposure..mo.+ Hearing_Level_3_Cat)
summary(plan_all_AoLE_HearStat)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     Age.of.Language.Exposure..mo. + Hearing_Level_3_Cat, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.7011 -2.4923 -0.6156  2.1823  8.8189 
## 
## Coefficients:
##                                    Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                        20.02871    2.29794   8.716  8.5e-14 ***
## SES..3.66.                         -0.06211    0.02692  -2.307  0.02322 *  
## SexMale                            -0.05949    0.70033  -0.085  0.93248    
## AgeMonths                          -0.04095    0.03173  -1.290  0.20001    
## Age.of.Language.Exposure..mo.       0.06796    0.02477   2.744  0.00725 ** 
## Hearing_Level_3_CatDeaf            -1.14683    0.93669  -1.224  0.22382    
## Hearing_Level_3_CatHard of Hearing -0.04387    1.19832  -0.037  0.97087    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.478 on 96 degrees of freedom
##   (6 observations deleted due to missingness)
## Multiple R-squared:  0.1545, Adjusted R-squared:  0.1017 
## F-statistic: 2.925 on 6 and 96 DF,  p-value: 0.01156

AIC(plan_all_AoLE_HearStat)

## [1] 557.8476

Compare Models from Step 3a-c to model from Step 2:

anova(plan_all_AoLE, plan_all_AoLE_AoAE)

## Analysis of Variance Table
## 
## Model 1: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1267.3                           
## 2    103 1246.7  1    20.647 1.7059 0.1944

anova(plan_all_AoLE, plan_all_AoLE_Modality)

## Analysis of Variance Table
## 
## Model 1: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Language_Modality
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1267.3                           
## 2    103 1234.6  1    32.679 2.7263 0.1018

anova(plan_all_AoLE, plan_all_AoLE_LoAE)

## Analysis of Variance Table
## 
## Model 1: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo.
## Model 2: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + Age.of.Language.Exposure..mo. + 
##     Length.of.Auditory.Exposure
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1267.3                           
## 2    103 1246.6  1    20.668 1.7077 0.1942

Checking assumptions for best-fit Plan/Organize model

ols_plot_resid_qq(plan_all_AoLE)

ols_test_normality(plan_all_AoLE)

## -----------------------------------------------
##        Test             Statistic       pvalue  
## -----------------------------------------------
## Shapiro-Wilk              0.979          0.0831 
## Kolmogorov-Smirnov        0.0865         0.3880 
## Cramer-von Mises          9.0793         0.0000 
## Anderson-Darling          0.7187         0.0591 
## -----------------------------------------------

ols_plot_resid_hist(plan_all_AoLE)

ols_test_correlation(plan_all_AoLE)

## [1] 0.9908092

ols_plot_resid_fit(plan_all_AoLE)

Table 4.

stargazer(overall_all_AoLE_AoAE, inhibition_all_AoLE_AoAE, shift_all_AoLE_AoAE, emotctrl_all_AoLE_AoAE, wm_all_AoLE_AoAE, plan_all_AoLE_AoAE, type= "html", title = "Linear Regression Results for Raw Scores", align=TRUE, dep.var.labels=c("Global Executive Composite", "Inhibition", "Shift", "Emotional Control", "Working Memory", "Planning/Organization"), covariate.labels=c("Socioeconomic Status (SES)", "Sex (Male)", "Age (Months)", "Age of Exposure to Language (Months)", "Age of Auditory Exposure (Months)"), out="Table4.html")

Question 3: Do children with later or degraded access to auditory and/or language input exhibit clinically-significant deficits in executive functioning skills relative to typically-hearing peers?

This analysis looks at T-scores, and therefore only includes participants who were within the normed age range for the BRIEF-P

Create dataframe with participants who are age appropriate for the BRIEF-P (less than 5 years 11 Months old)

BRIEF_AgeApp <- subset(BRIEF, BRIEF$AgeMonths <= 71)

This is the dataframe used in Question 3 Analyses
98 participants

Demographics for the Question 3 subset

table1::label(BRIEF_AgeApp$AgeMonths) <- "Age (Months)"
table1::label(BRIEF_AgeApp$SES..3.66.) <- "SES"
table1::label(BRIEF_AgeApp$Sex) <- "Sex"
table1::label(BRIEF_AgeApp$Race_recoded) <- "Race"
table1::label(BRIEF_AgeApp$Ethnicity_recoded) <- "Ethnicity"
table1::label(BRIEF_AgeApp$Age.of.Language.Exposure..mo.) <- "Age of Language Exposure (Months)"
table1::label(BRIEF_AgeApp$AoAE) <- "Age of Auditory Exposure (Months)"

 
table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded + Age.of.Language.Exposure..mo. + AoAE, data = BRIEF_AgeApp)

	Overall (N=98)
Age (Months)
Mean (SD)	55.2 (9.87)
Median [Min, Max]	56.0 [37.0, 71.0]
SES
Mean (SD)	51.3 (12.3)
Median [Min, Max]	53.5 [9.00, 66.0]
Sex
Female	51 (52.0%)
Male	47 (48.0%)
Race
Asian	1 (1.0%)
Black or African American	1 (1.0%)
White	87 (88.8%)
More than one	7 (7.1%)
Other/Missing	2 (2.0%)
Ethnicity
Hispanic	6 (6.1%)
Non-Hispanic	78 (79.6%)
Prefer not to answer	1 (1.0%)
Missing	13 (13.3%)
Age of Language Exposure (Months)
Mean (SD)	9.17 (15.7)
Median [Min, Max]	0 [0, 58.0]
Missing	1 (1.0%)
Age of Auditory Exposure (Months)
Mean (SD)	15.1 (21.0)
Median [Min, Max]	0.500 [0, 70.0]
Missing	8 (8.2%)

Comparing demographics by Language group (within Question 3 subset)

table1::table1(~AgeMonths + SES..3.66. + Sex + Race_recoded + Ethnicity_recoded + Age.of.Language.Exposure..mo. + AoAE | LanguageGroup, data = BRIEF_AgeApp, overall=F)

	Typically Hearing (N=44)	Early ASL (N=20)	Later English (N=20)	Later ASL (N=14)
Age (Months)
Mean (SD)	53.7 (9.80)	54.5 (8.98)	58.3 (9.67)	56.8 (11.3)
Median [Min, Max]	54.0 [37.0, 71.0]	54.5 [41.0, 71.0]	60.0 [37.0, 71.0]	60.0 [37.0, 71.0]
SES
Mean (SD)	55.5 (8.83)	47.0 (16.0)	49.6 (11.3)	46.7 (14.2)
Median [Min, Max]	56.0 [21.5, 66.0]	53.8 [19.0, 66.0]	50.5 [15.0, 66.0]	50.8 [9.00, 62.0]
Sex
Female	23 (52.3%)	11 (55.0%)	10 (50.0%)	7 (50.0%)
Male	21 (47.7%)	9 (45.0%)	10 (50.0%)	7 (50.0%)
Race
Asian	0 (0%)	0 (0%)	0 (0%)	1 (7.1%)
Black or African American	0 (0%)	0 (0%)	1 (5.0%)	0 (0%)
White	41 (93.2%)	18 (90.0%)	17 (85.0%)	11 (78.6%)
More than one	3 (6.8%)	1 (5.0%)	2 (10.0%)	1 (7.1%)
Other/Missing	0 (0%)	1 (5.0%)	0 (0%)	1 (7.1%)
Ethnicity
Hispanic	3 (6.8%)	0 (0%)	1 (5.0%)	2 (14.3%)
Non-Hispanic	40 (90.9%)	13 (65.0%)	15 (75.0%)	10 (71.4%)
Prefer not to answer	0 (0%)	1 (5.0%)	0 (0%)	0 (0%)
Missing	1 (2.3%)	6 (30.0%)	4 (20.0%)	2 (14.3%)
Age of Language Exposure (Months)
Mean (SD)	0 (0)	0 (0)	22.0 (17.2)	34.6 (8.23)
Median [Min, Max]	0 [0, 0]	0 [0, 0]	20.0 [0.500, 58.0]	36.0 [18.0, 46.0]
Missing	0 (0%)	0 (0%)	0 (0%)	1 (7.1%)
Age of Auditory Exposure (Months)
Mean (SD)	0 (0)	44.4 (20.9)	22.0 (17.2)	25.2 (18.3)
Median [Min, Max]	0 [0, 0]	46.5 [1.00, 70.0]	20.0 [0, 58.0]	27.0 [3.00, 60.0]
Missing	0 (0%)	6 (30.0%)	0 (0%)	2 (14.3%)

Figure 1.

G <- ggplot(data=BRIEF_AgeApp, mapping = aes(x= Language_Timing, y=GEC_Tscore)) + geom_boxplot() + geom_dotplot(mapping=aes(fill=Language_Modality), dotsize = 1.1, method = "dotdensity", binaxis = "y", stackdir = "center", alpha=0.8) + theme(text = element_text(size=20), axis.title=element_text(size=20,face="bold"), legend.title=element_text(size=24), legend.text=element_text(size=24), axis.title.x=element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) + ylab("Global Executive Composite (T-scores)") + scale_x_discrete(labels=c("Early" = "Early Language", "Later" = "Later Language")) + scale_y_continuous(limits= c(30, 100), breaks= c(30, 40, 50, 60, 70, 80, 90, 100)) + scale_fill_grey(name = "Language Modality", start=0.1, end=0.9, ) + geom_hline(aes(yintercept= 50, linetype = "Mean"), colour= 'black') + geom_hline(aes(yintercept= 65, linetype = "Elevated"), colour= 'grey17') + scale_linetype_manual(name = "Standard Score", values=c("dotted", "longdash"), guide = guide_legend(override.aes = list(color = c("grey17", "black"))))

I <- ggplot(data=BRIEF_AgeApp, mapping = aes(x= Language_Timing, y=Inhibit_Tscore)) + geom_boxplot() + geom_dotplot(mapping=aes(fill=Language_Modality), dotsize = 1.1, method = "dotdensity", binaxis = "y", stackdir = "center", alpha=0.8) + theme(text = element_text(size=20), axis.title=element_text(size=20,face="bold"), legend.title=element_text(size=24), legend.text=element_text(size=24), axis.title.x=element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) + ylab("Inhibition (T-scores)") + scale_x_discrete(labels=c("Early" = "Early Language", "Later" = "Later Language")) + scale_y_continuous(limits= c(30, 100), breaks= c(30, 40, 50, 60, 70, 80, 90, 100)) + scale_fill_grey(name = "Language Modality", start=0.1, end=0.9) + geom_hline(aes(yintercept= 50, linetype = "Mean"), colour= 'black') + geom_hline(aes(yintercept= 65, linetype = "Elevated"), colour= 'grey17') + scale_linetype_manual(name = "Standard Score", values=c("dotted", "longdash"), guide = guide_legend(override.aes = list(color = c("grey17", "black"))))

S <- ggplot(data=BRIEF_AgeApp, mapping = aes(x= Language_Timing, y=Shift_Tscore)) + geom_boxplot() + geom_dotplot(mapping=aes(fill=Language_Modality), dotsize = 1.1, method = "dotdensity", binaxis = "y", stackdir = "center", alpha=0.8) + theme(text = element_text(size=20), axis.title=element_text(size=20,face="bold"), legend.title=element_text(size=24), legend.text=element_text(size=24), axis.title.x=element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) + ylab("Shift (T-scores)") + scale_x_discrete(labels=c("Early" = "Early Language", "Later" = "Later Language")) + scale_y_continuous(limits= c(30, 100), breaks= c(30, 40, 50, 60, 70, 80, 90, 100)) + scale_fill_grey(name = "Language Modality", start=0.1, end=0.9) + geom_hline(aes(yintercept= 50, linetype = "Mean"), colour= 'black') + geom_hline(aes(yintercept= 65, linetype = "Elevated"), colour= 'grey17') + scale_linetype_manual(name = "Standard Score", values=c("dotted", "longdash"), guide = guide_legend(override.aes = list(color = c("grey17", "black"))))

E <- ggplot(data=BRIEF_AgeApp, mapping = aes(x= Language_Timing, y= Emotional.Control_Tscore)) + geom_boxplot() + geom_dotplot(mapping=aes(fill=Language_Modality), dotsize = 1.1, method = "dotdensity", binaxis = "y", stackdir = "center", alpha=0.8) + theme(text = element_text(size=20), axis.title=element_text(size=20,face="bold"), legend.title=element_text(size=24), legend.text=element_text(size=24), axis.title.x=element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) + ylab("Emotional Control (T-scores)") + scale_x_discrete(labels=c("Early" = "Early Language", "Later" = "Later Language")) + scale_y_continuous(limits= c(30, 100), breaks= c(30, 40, 50, 60, 70, 80, 90, 100)) + scale_fill_grey(name = "Language Modality", start=0.1, end=0.9)  + geom_hline(aes(yintercept= 50, linetype = "Mean"), colour= 'black') + geom_hline(aes(yintercept= 65, linetype = "Elevated"), colour= 'grey17') + scale_linetype_manual(name = "Standard Score", values=c("dotted", "longdash"), guide = guide_legend(override.aes = list(color = c("grey17", "black"))))

W <- ggplot(data=BRIEF_AgeApp, mapping = aes(x= Language_Timing, y= Working.Memory_Tscore)) + geom_boxplot() + geom_dotplot(mapping=aes(fill=Language_Modality), dotsize = 1.1, method = "dotdensity", binaxis = "y", stackdir = "center", alpha=0.8) + theme(text = element_text(size=20), axis.title=element_text(size=20,face="bold"), legend.title=element_text(size=24), legend.text=element_text(size=24), axis.title.x=element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black"))  + ylab("Working Memory (T-scores)") + scale_x_discrete(labels=c("Early" = "Early Language", "Later" = "Later Language")) + scale_y_continuous(limits= c(30, 100), breaks= c(30, 40, 50, 60, 70, 80, 90, 100)) + scale_fill_grey(name = "Language Modality", start=0.1, end=0.9)  + geom_hline(aes(yintercept= 50, linetype = "Mean"), colour= 'black') + geom_hline(aes(yintercept= 65, linetype = "Elevated"), colour= 'grey17') + scale_linetype_manual(name = "Standard Score", values=c("dotted", "longdash"), guide = guide_legend(override.aes = list(color = c("grey17", "black"))))

P <- ggplot(data=BRIEF_AgeApp, mapping = aes(x= Language_Timing, y= Plan.Organize_Tscore)) + geom_boxplot() + geom_dotplot(mapping=aes(fill=Language_Modality), dotsize = 1.1, method = "dotdensity", binaxis = "y", stackdir = "center", alpha=0.8) + theme(text = element_text(size=20), axis.title=element_text(size=20,face="bold"), legend.title=element_text(size=24), legend.text=element_text(size=24), axis.title.x=element_blank(), panel.grid.major = element_blank(), panel.grid.minor = element_blank(), panel.background = element_blank(), axis.line = element_line(colour = "black")) + ylab("Plan/Organize (T-scores)") + scale_x_discrete(labels=c("Early" = "Early Language", "Later" = "Later Language")) + scale_y_continuous(limits= c(30, 100), breaks= c(30, 40, 50, 60, 70, 80, 90, 100)) + scale_fill_grey(name = "Language Modality", start=0.1, end=0.9) + geom_hline(aes(yintercept= 50, linetype = "Mean"), colour= 'black') + geom_hline(aes(yintercept= 65, linetype = "Elevated"), colour= 'grey17') + scale_linetype_manual(name = "Standard Score", values=c("dotted", "longdash"), guide = guide_legend(override.aes = list(color = c("grey17", "black"))))

ggarrange(G, I, S, E, W, P, ncol=3, nrow=2, common.legend = TRUE, legend="bottom")

Table 5 data & analyses

Risk Ratios (Relative Risk) and 95% confidence intervals for Global Executive Composite T-scores

BRIEF_AgeApp_GEC_over65 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Global Executive Composite T-score < 65 (n)" = sum(GEC_Tscore < 65), "Global Executive Composite T-score > 65 (n)" = sum(GEC_Tscore >= 65))

BRIEF_AgeApp_GEC_over65 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Global Executive Composite T-score < 65 (n)	Global Executive Composite T-score > 65 (n)
Typically Hearing	41	3
Early ASL	19	1
Later English	17	3
Later ASL	12	2

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
GEC_rr <- matrix(c(41, 3, 19, 1, 17, 3, 12, 2), 4, 2, byrow=TRUE)
dimnames(GEC_rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "GEC T-score" = c("Not Clinically Significant", "Clinically Significant"))
GEC_rr

##                    GEC T-score
## Group               Not Clinically Significant Clinically Significant
##   Typically Hearing                         41                      3
##   Early ASL                                 19                      1
##   Later English                             17                      3
##   Later ASL                                 12                      2

riskratio.small(GEC_rr, verbose=TRUE)

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## $x
##                    GEC T-score
## Group               Not Clinically Significant Clinically Significant
##   Typically Hearing                         41                      3
##   Early ASL                                 19                      1
##   Later English                             17                      3
##   Later ASL                                 12                      2
## 
## $data
##                    GEC T-score
## Group               Not Clinically Significant Clinically Significant Total
##   Typically Hearing                         41                      3    44
##   Early ASL                                 19                      1    20
##   Later English                             17                      3    20
##   Later ASL                                 12                      2    14
##   Total                                     89                      9    98
## 
## $p.exposed
##                    GEC T-score
## Group               Not Clinically Significant Clinically Significant     Total
##   Typically Hearing                  0.4606742              0.3333333 0.4489796
##   Early ASL                          0.2134831              0.1111111 0.2040816
##   Later English                      0.1910112              0.3333333 0.2040816
##   Later ASL                          0.1348315              0.2222222 0.1428571
##   Total                              1.0000000              1.0000000 1.0000000
## 
## $p.outcome
##                    GEC T-score
## Group               Not Clinically Significant Clinically Significant Total
##   Typically Hearing                  0.9318182             0.06818182     1
##   Early ASL                          0.9500000             0.05000000     1
##   Later English                      0.8500000             0.15000000     1
##   Later ASL                          0.8571429             0.14285714     1
##   Total                              0.9081633             0.09183673     1
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate      lower    upper
##   Typically Hearing 1.000000         NA       NA
##   Early ASL         0.562500 0.06229044 5.079531
##   Later English     1.687500 0.37260128 7.642637
##   Later ASL         1.607143 0.29800962 8.667197
## 
## $conf.level
## [1] 0.95
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.8441962    1.0000000  0.7806100
##   Later English      0.3428741    0.3662788  0.2979414
##   Later ASL          0.4334207    0.5852322  0.3858989
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

Inhibit Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_Inhibit_over65 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Inhibit T-score < 65 (n)" = sum(Inhibit_Tscore < 65), "Inhibit T-score > 65 (n)" = sum(Inhibit_Tscore >= 65))

BRIEF_AgeApp_Inhibit_over65 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Inhibit T-score < 65 (n)	Inhibit T-score > 65 (n)
Typically Hearing	42	2
Early ASL	18	2
Later English	19	1
Later ASL	14	0

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
Inhibit_rr <- matrix(c(42,2, 18, 2, 19, 1, 14, 0), 4, 2, byrow=TRUE)
dimnames(Inhibit_rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "GEC T-score" = c("Not Clinically Significant", "Clinically Significant"))
Inhibit_rr

##                    GEC T-score
## Group               Not Clinically Significant Clinically Significant
##   Typically Hearing                         42                      2
##   Early ASL                                 18                      2
##   Later English                             19                      1
##   Later ASL                                 14                      0

riskratio.small(Inhibit_rr)

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## $data
##                    GEC T-score
## Group               Not Clinically Significant Clinically Significant Total
##   Typically Hearing                         42                      2    44
##   Early ASL                                 18                      2    20
##   Later English                             19                      1    20
##   Later ASL                                 14                      0    14
##   Total                                     93                      5    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate      lower    upper
##   Typically Hearing     1.00         NA       NA
##   Early ASL             1.50 0.22720962 9.902750
##   Later English         0.75 0.07213456 7.797927
##   Later ASL             0.00 0.00000000      NaN
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.4560292     0.583113  0.4033953
##   Later English      0.9101382     1.000000  0.9364430
##   Later ASL          0.5722928     1.000000  0.4168811
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

Shift Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_Shift_over65 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Shift T-score < 65 (n)" = sum(Shift_Tscore < 65), "Shift T-score > 65 (n)" = sum(Shift_Tscore >= 65))

BRIEF_AgeApp_Shift_over65 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Shift T-score < 65 (n)	Shift T-score > 65 (n)
Typically Hearing	42	2
Early ASL	19	1
Later English	18	2
Later ASL	13	1

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
Shift_rr <- matrix(c(42,2,19,1, 18, 2, 13, 1), 4, 2, byrow=TRUE)
dimnames(Shift_rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Shift T-score" = c("Not Clinically Significant", "Clinically Significant"))
Shift_rr

##                    Shift T-score
## Group               Not Clinically Significant Clinically Significant
##   Typically Hearing                         42                      2
##   Early ASL                                 19                      1
##   Later English                             18                      2
##   Later ASL                                 13                      1

riskratio.small(Shift_rr)

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## $data
##                    Shift T-score
## Group               Not Clinically Significant Clinically Significant Total
##   Typically Hearing                         42                      2    44
##   Early ASL                                 19                      1    20
##   Later English                             18                      2    20
##   Later ASL                                 13                      1    14
##   Total                                     92                      6    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate      lower     upper
##   Typically Hearing 1.000000         NA        NA
##   Early ASL         0.750000 0.07213456  7.797927
##   Later English     1.500000 0.22720962  9.902750
##   Later ASL         1.071429 0.10488371 10.945067
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.9101382     1.000000  0.9364430
##   Later English      0.4560292     0.583113  0.4033953
##   Later ASL          0.7123412     1.000000  0.7023059
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

Emotional Control Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_EmoCtrl_over65 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Emotional Control T-score < 65 (n)" = sum(Emotional.Control_Tscore < 65), "Emotional Control T-score > 65 (n)" = sum(Emotional.Control_Tscore >= 65))

BRIEF_AgeApp_EmoCtrl_over65 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Emotional Control T-score < 65 (n)	Emotional Control T-score > 65 (n)
Typically Hearing	41	3
Early ASL	20	0
Later English	18	2
Later ASL	12	2

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
EmoCtrl_rr <- matrix(c(41,3,20,0, 18, 2, 12, 2), 4, 2, byrow=TRUE)
dimnames(EmoCtrl_rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Emotional Control T-score" = c("Not Clinically Significant", "Clinically Significant"))
EmoCtrl_rr

##                    Emotional Control T-score
## Group               Not Clinically Significant Clinically Significant
##   Typically Hearing                         41                      3
##   Early ASL                                 20                      0
##   Later English                             18                      2
##   Later ASL                                 12                      2

riskratio.small(EmoCtrl_rr)

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## $data
##                    Emotional Control T-score
## Group               Not Clinically Significant Clinically Significant Total
##   Typically Hearing                         41                      3    44
##   Early ASL                                 20                      0    20
##   Later English                             18                      2    20
##   Later ASL                                 12                      2    14
##   Total                                     91                      7    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate     lower    upper
##   Typically Hearing 1.000000        NA       NA
##   Early ASL         0.000000 0.0000000      NaN
##   Later English     1.125000 0.2036087 6.215966
##   Later ASL         1.607143 0.2980096 8.667197
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.3178763    0.5458909  0.2316502
##   Later English      0.6729097    0.6439090  0.6601990
##   Later ASL          0.4334207    0.5852322  0.3858989
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

Working Memory Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_WorkMem_over65 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Working Memory T-score < 65 (n)" = sum(Working.Memory_Tscore < 65), "Working Memory T-score > 65 (n)" = sum(Working.Memory_Tscore >= 65))

BRIEF_AgeApp_WorkMem_over65 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Working Memory T-score < 65 (n)	Working Memory T-score > 65 (n)
Typically Hearing	39	5
Early ASL	17	3
Later English	18	2
Later ASL	13	1

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
WorkMem_rr <- matrix(c(39,5,17,3, 18, 2, 13, 1), 4, 2, byrow=TRUE)
dimnames(WorkMem_rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Working Memory T-score" = c("Not Clinically Significant", "Clinically Significant"))
WorkMem_rr

##                    Working Memory T-score
## Group               Not Clinically Significant Clinically Significant
##   Typically Hearing                         39                      5
##   Early ASL                                 17                      3
##   Later English                             18                      2
##   Later ASL                                 13                      1

riskratio.small(WorkMem_rr)

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## $data
##                    Working Memory T-score
## Group               Not Clinically Significant Clinically Significant Total
##   Typically Hearing                         39                      5    44
##   Early ASL                                 17                      3    20
##   Later English                             18                      2    20
##   Later ASL                                 13                      1    14
##   Total                                     87                     11    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group                estimate      lower    upper
##   Typically Hearing 1.0000000         NA       NA
##   Early ASL         1.1250000 0.29748109 4.254472
##   Later English     0.7500000 0.15881991 3.541747
##   Later ASL         0.5357143 0.06820514 4.207744
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.6878536    0.6969793  0.6834809
##   Later English      0.9100612    1.0000000  0.8712975
##   Later ASL          0.7244461    1.0000000  0.6515079
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

Plan/Organize Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_PlanOrg_over65 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Plan/Organize T-score < 65 (n)" = sum(Plan.Organize_Tscore < 65), "Plan/Organize T-score > 65 (n)" = sum(Plan.Organize_Tscore >= 65))

BRIEF_AgeApp_PlanOrg_over65 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Plan/Organize T-score < 65 (n)	Plan/Organize T-score > 65 (n)
Typically Hearing	41	3
Early ASL	19	1
Later English	16	4
Later ASL	14	0

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
PlanOrg_rr <- matrix(c(41,3,19,1, 16, 4, 14, 0), 4, 2, byrow=TRUE)
dimnames(PlanOrg_rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Plan/Organize T-score" = c("Not Clinically Significant", "Clinically Significant"))
PlanOrg_rr

##                    Plan/Organize T-score
## Group               Not Clinically Significant Clinically Significant
##   Typically Hearing                         41                      3
##   Early ASL                                 19                      1
##   Later English                             16                      4
##   Later ASL                                 14                      0

riskratio.small(PlanOrg_rr)

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## Warning in chisq.test(xx, correct = correction): Chi-squared approximation may
## be incorrect

## $data
##                    Plan/Organize T-score
## Group               Not Clinically Significant Clinically Significant Total
##   Typically Hearing                         41                      3    44
##   Early ASL                                 19                      1    20
##   Later English                             16                      4    20
##   Later ASL                                 14                      0    14
##   Total                                     90                      8    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate      lower    upper
##   Typically Hearing   1.0000         NA       NA
##   Early ASL           0.5625 0.06229044 5.079531
##   Later English       2.2500 0.55455158 9.128998
##   Later ASL           0.0000 0.00000000      NaN
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.8441962    1.0000000  0.7806100
##   Later English      0.1562526    0.1914591  0.1173246
##   Later ASL          0.4292196    1.0000000  0.3157160
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

Additional checks

Check 1: Comparing stepwise models with AoAE added first

Possible Issue: Because Age of Language Exposure and Age of Auditory Exposure were significantly correlated, we need to determine whether they might actually be measuring same underlying construct

Approach: See whether findings above (where Language but not Auditory exposure significantly predicts BRIEF scores) hold when predictors are added to the model in a different order

overall_all_base <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths)
summary(overall_all_base)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -32.330 -14.855  -0.360   8.628  60.531 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 107.1580    11.5903   9.245 3.02e-15 ***
## SES..3.66.   -0.4345     0.1362  -3.191  0.00187 ** 
## SexMale      -1.1793     3.6935  -0.319  0.75014    
## AgeMonths     0.1167     0.1420   0.822  0.41320    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.94 on 105 degrees of freedom
## Multiple R-squared:  0.09454,    Adjusted R-squared:  0.06867 
## F-statistic: 3.654 on 3 and 105 DF,  p-value: 0.01492

overall_all_AoAE <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths+AoAE)
summary(overall_all_AoAE)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -32.739 -15.336  -0.297   8.446  60.008 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 107.00901   11.66195   9.176 4.66e-15 ***
## SES..3.66.   -0.44157    0.14033  -3.147  0.00215 ** 
## SexMale      -1.13741    3.71503  -0.306  0.76009    
## AgeMonths     0.13112    0.15648   0.838  0.40400    
## AoAE         -0.01834    0.08166  -0.225  0.82276    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 19.03 on 104 degrees of freedom
## Multiple R-squared:  0.09498,    Adjusted R-squared:  0.06017 
## F-statistic: 2.729 on 4 and 104 DF,  p-value: 0.03311

anova(overall_all_base, overall_all_AoAE)

## Analysis of Variance Table
## 
## Model 1: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
##   Res.Df   RSS Df Sum of Sq      F Pr(>F)
## 1    105 37686                           
## 2    104 37668  1    18.264 0.0504 0.8228

overall_all_AoAE_AoLE <- lm(data=BRIEF_AgeOf, formula = GEC_RawScore~SES..3.66.+Sex+AgeMonths+AoAE+Age.of.Language.Exposure..mo.)
summary(overall_all_AoAE_AoLE)

## 
## Call:
## lm(formula = GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE + 
##     Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -35.488 -13.012  -0.955   9.269  64.745 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   109.30858   11.24579   9.720 3.11e-16 ***
## SES..3.66.                     -0.35417    0.13801  -2.566  0.01172 *  
## SexMale                        -1.70684    3.57928  -0.477  0.63447    
## AgeMonths                      -0.02933    0.15945  -0.184  0.85440    
## AoAE                           -0.08470    0.08152  -1.039  0.30121    
## Age.of.Language.Exposure..mo.   0.34130    0.11166   3.057  0.00285 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 18.31 on 103 degrees of freedom
## Multiple R-squared:  0.1702, Adjusted R-squared:   0.13 
## F-statistic: 4.226 on 5 and 103 DF,  p-value: 0.001553

anova(overall_all_AoAE, overall_all_AoAE_AoLE)

## Analysis of Variance Table
## 
## Model 1: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
## Model 2: GEC_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE + Age.of.Language.Exposure..mo.
##   Res.Df   RSS Df Sum of Sq      F   Pr(>F)   
## 1    104 37668                                
## 2    103 34535  1    3132.4 9.3425 0.002852 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

#~~~~~#
inhibition_all_base <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths)
summary(inhibition_all_base)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.4961  -4.5032   0.0108   4.0216  21.5107 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 25.07731    3.65405   6.863 4.86e-10 ***
## SES..3.66.  -0.10990    0.04292  -2.560   0.0119 *  
## SexMale      0.97828    1.16445   0.840   0.4028    
## AgeMonths    0.06696    0.04478   1.495   0.1378    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.973 on 105 degrees of freedom
## Multiple R-squared:  0.08539,    Adjusted R-squared:  0.05926 
## F-statistic: 3.268 on 3 and 105 DF,  p-value: 0.02424

inhibition_all_AoAE <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths+AoAE)
summary(inhibition_all_AoAE)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.2413  -4.6083   0.1954   4.1177  21.2498 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 25.002918   3.675294   6.803  6.7e-10 ***
## SES..3.66.  -0.113419   0.044224  -2.565   0.0118 *  
## SexMale      0.999205   1.170801   0.853   0.3954    
## AgeMonths    0.074163   0.049315   1.504   0.1357    
## AoAE        -0.009156   0.025737  -0.356   0.7227    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.998 on 104 degrees of freedom
## Multiple R-squared:  0.0865, Adjusted R-squared:  0.05137 
## F-statistic: 2.462 on 4 and 104 DF,  p-value: 0.04977

anova(inhibition_all_base, inhibition_all_AoAE)

## Analysis of Variance Table
## 
## Model 1: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    105 3745.7                           
## 2    104 3741.2  1    4.5532 0.1266 0.7227

inhibition_all_AoAE_AoLE <- lm(data=BRIEF_AgeOf, formula = Inhibit_RawScore~SES..3.66.+Sex+AgeMonths+AoAE+Age.of.Language.Exposure..mo.)
summary(inhibition_all_AoAE_AoLE)

## 
## Call:
## lm(formula = Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE + Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##      Min       1Q   Median       3Q      Max 
## -11.3417  -4.3176  -0.1074   4.1316  22.2424 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.48478    3.63271   7.015 2.49e-10 ***
## SES..3.66.                    -0.09511    0.04458  -2.133   0.0353 *  
## SexMale                        0.87988    1.15621   0.761   0.4484    
## AgeMonths                      0.04054    0.05151   0.787   0.4330    
## AoAE                          -0.02306    0.02633  -0.876   0.3832    
## Age.of.Language.Exposure..mo.  0.07152    0.03607   1.983   0.0501 .  
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.915 on 103 degrees of freedom
## Multiple R-squared:  0.1201, Adjusted R-squared:  0.07737 
## F-statistic: 2.811 on 5 and 103 DF,  p-value: 0.02018

anova(inhibition_all_AoAE, inhibition_all_AoAE_AoLE)

## Analysis of Variance Table
## 
## Model 1: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
## Model 2: Inhibit_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE + Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F  Pr(>F)  
## 1    104 3741.2                              
## 2    103 3603.6  1    137.54 3.9313 0.05006 .
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

#~~~~~#
shift_all_base <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths) 
summary(shift_all_base)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths, data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -4.936 -2.682 -0.776  1.774 14.985 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 17.93727    2.20373   8.140 8.65e-13 ***
## SES..3.66.  -0.08380    0.02589  -3.237  0.00162 ** 
## SexMale     -0.79353    0.70227  -1.130  0.26107    
## AgeMonths    0.01077    0.02701   0.399  0.69088    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.602 on 105 degrees of freedom
## Multiple R-squared:  0.0979, Adjusted R-squared:  0.07213 
## F-statistic: 3.799 on 3 and 105 DF,  p-value: 0.01245

shift_all_AoAE <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths+AoAE)  
summary(shift_all_AoAE)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -4.7838 -2.6726 -0.7207  1.9186 14.8359 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 18.012293   2.214116   8.135 9.32e-13 ***
## SES..3.66.  -0.080255   0.026642  -3.012  0.00326 ** 
## SexMale     -0.814632   0.705328  -1.155  0.25075    
## AgeMonths    0.003507   0.029709   0.118  0.90625    
## AoAE         0.009233   0.015505   0.596  0.55279    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.613 on 104 degrees of freedom
## Multiple R-squared:  0.101,  Adjusted R-squared:  0.06639 
## F-statistic:  2.92 on 4 and 104 DF,  p-value: 0.02467

anova(shift_all_base, shift_all_AoAE)

## Analysis of Variance Table
## 
## Model 1: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    105 1362.4                           
## 2    104 1357.8  1      4.63 0.3546 0.5528

shift_all_AoAE_AoLE <- lm(data=BRIEF_AgeOf, formula = Shift_RawScore~SES..3.66.+Sex+AgeMonths+AoAE+Age.of.Language.Exposure..mo.)  
summary(shift_all_AoAE_AoLE)

## 
## Call:
## lm(formula = Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE + Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.4369 -2.2772 -0.5143  1.4576 14.0693 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   18.425943   2.144990   8.590 9.89e-14 ***
## SES..3.66.                    -0.064533   0.026323  -2.452   0.0159 *  
## SexMale                       -0.917061   0.682702  -1.343   0.1821    
## AgeMonths                     -0.025355   0.030413  -0.834   0.4064    
## AoAE                          -0.002704   0.015549  -0.174   0.8623    
## Age.of.Language.Exposure..mo.  0.061394   0.021298   2.883   0.0048 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.493 on 103 degrees of freedom
## Multiple R-squared:  0.1681, Adjusted R-squared:  0.1277 
## F-statistic: 4.162 on 5 and 103 DF,  p-value: 0.001745

anova(shift_all_AoAE, shift_all_AoAE_AoLE)

## Analysis of Variance Table
## 
## Model 1: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
## Model 2: Shift_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE + Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F   Pr(>F)   
## 1    104 1357.8                                
## 2    103 1256.4  1    101.36 8.3093 0.004802 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

#~~~~~#
emotctrl_all_base <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths)
emotctrl_all_AoAE <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths+AoAE)
summary(emotctrl_all_AoAE)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths + AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -8.6343 -3.0275 -0.6445  2.5623 13.9640 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 19.986525   2.512209   7.956  2.3e-12 ***
## SES..3.66.  -0.094341   0.030229  -3.121  0.00233 ** 
## SexMale     -0.688072   0.800289  -0.860  0.39189    
## AgeMonths    0.007952   0.033709   0.236  0.81396    
## AoAE        -0.007747   0.017592  -0.440  0.66059    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.1 on 104 degrees of freedom
## Multiple R-squared:  0.08881,    Adjusted R-squared:  0.05376 
## F-statistic: 2.534 on 4 and 104 DF,  p-value: 0.04459

anova(emotctrl_all_base, emotctrl_all_AoAE)

## Analysis of Variance Table
## 
## Model 1: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    105 1751.2                           
## 2    104 1748.0  1    3.2593 0.1939 0.6606

emotctrl_all_AoAE_AoLE <- lm(data=BRIEF_AgeOf, formula = Emotional.Control_RawScore~SES..3.66.+Sex+AgeMonths+AoAE+Age.of.Language.Exposure..mo.)
summary(emotctrl_all_AoAE_AoLE)

## 
## Call:
## lm(formula = Emotional.Control_RawScore ~ SES..3.66. + Sex + 
##     AgeMonths + AoAE + Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -8.418 -3.282 -0.533  2.551 13.559 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   20.204803   2.509534   8.051  1.5e-12 ***
## SES..3.66.                    -0.086045   0.030797  -2.794  0.00621 ** 
## SexMale                       -0.742122   0.798728  -0.929  0.35499    
## AgeMonths                     -0.007278   0.035581  -0.205  0.83833    
## AoAE                          -0.014046   0.018191  -0.772  0.44179    
## Age.of.Language.Exposure..mo.  0.032397   0.024918   1.300  0.19645    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 4.086 on 103 degrees of freedom
## Multiple R-squared:  0.1035, Adjusted R-squared:   0.06 
## F-statistic: 2.379 on 5 and 103 DF,  p-value: 0.0437

anova(emotctrl_all_AoAE, emotctrl_all_AoAE_AoLE)

## Analysis of Variance Table
## 
## Model 1: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
## Model 2: Emotional.Control_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE + 
##     Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    104 1748.0                           
## 2    103 1719.8  1    28.224 1.6904 0.1965

#~~~~~#
wm_all_base <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths)
summary(wm_all_base)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -8.748 -4.607 -1.414  3.432 23.472 
## 
## Coefficients:
##             Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 25.20501    3.70291   6.807 6.38e-10 ***
## SES..3.66.  -0.07306    0.04350  -1.680    0.096 .  
## SexMale     -0.43333    1.18002  -0.367    0.714    
## AgeMonths    0.04043    0.04538   0.891    0.375    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 6.053 on 105 degrees of freedom
## Multiple R-squared:  0.0346, Adjusted R-squared:  0.007015 
## F-statistic: 1.254 on 3 and 105 DF,  p-value: 0.294

wm_all_AoAE <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths+AoAE)
summary(wm_all_AoAE)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -8.664 -4.636 -1.421  3.421 23.385 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 25.180342   3.726467   6.757 8.35e-10 ***
## SES..3.66.  -0.074223   0.044840  -1.655    0.101    
## SexMale     -0.426386   1.187103  -0.359    0.720    
## AgeMonths    0.042816   0.050002   0.856    0.394    
## AoAE        -0.003036   0.026095  -0.116    0.908    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 6.081 on 104 degrees of freedom
## Multiple R-squared:  0.03472,    Adjusted R-squared:  -0.002403 
## F-statistic: 0.9353 on 4 and 104 DF,  p-value: 0.4466

anova(wm_all_base, wm_all_AoAE)

## Analysis of Variance Table
## 
## Model 1: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
##   Res.Df    RSS Df Sum of Sq      F Pr(>F)
## 1    105 3846.6                           
## 2    104 3846.1  1   0.50068 0.0135 0.9076

wm_all_AoAE_AoLE <- lm(data=BRIEF_AgeOf, formula = Working.Memory_RawScore~SES..3.66.+Sex+AgeMonths+AoAE+Age.of.Language.Exposure..mo.)
summary(wm_all_AoAE_AoLE)

## 
## Call:
## lm(formula = Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE + Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##    Min     1Q Median     3Q    Max 
## -9.313 -4.115 -1.080  3.208 24.929 
## 
## Coefficients:
##                                Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   25.929624   3.586996   7.229 8.82e-11 ***
## SES..3.66.                    -0.045745   0.044020  -1.039  0.30115    
## SexMale                       -0.611926   1.141660  -0.536  0.59312    
## AgeMonths                     -0.009465   0.050858  -0.186  0.85272    
## AoAE                          -0.024660   0.026002  -0.948  0.34514    
## Age.of.Language.Exposure..mo.  0.111208   0.035616   3.122  0.00233 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 5.841 on 103 degrees of freedom
## Multiple R-squared:  0.1182, Adjusted R-squared:  0.07538 
## F-statistic: 2.761 on 5 and 103 DF,  p-value: 0.02209

anova(wm_all_AoAE, wm_all_AoAE_AoLE)

## Analysis of Variance Table
## 
## Model 1: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
## Model 2: Working.Memory_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE + 
##     Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F   Pr(>F)   
## 1    104 3846.1                                
## 2    103 3513.5  1    332.57 9.7494 0.002329 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

#~~~~~#
plan_all_base <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths)
summary(plan_all_base)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths, 
##     data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.2655 -2.8630 -0.4688  2.5755  8.6681 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 18.888939   2.203955   8.570 9.66e-14 ***
## SES..3.66.  -0.076399   0.025890  -2.951  0.00391 ** 
## SexMale     -0.224973   0.702342  -0.320  0.74936    
## AgeMonths   -0.003322   0.027010  -0.123  0.90234    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.602 on 105 degrees of freedom
## Multiple R-squared:  0.07663,    Adjusted R-squared:  0.05024 
## F-statistic: 2.904 on 3 and 105 DF,  p-value: 0.03826

plan_all_AoAE <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths+AoAE)
summary(plan_all_AoAE)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE, data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.2075 -2.9021 -0.5068  2.7758  8.5542 
## 
## Coefficients:
##              Estimate Std. Error t value Pr(>|t|)    
## (Intercept) 18.826930   2.215542   8.498 1.49e-13 ***
## SES..3.66.  -0.079332   0.026659  -2.976  0.00363 ** 
## SexMale     -0.207528   0.705783  -0.294  0.76931    
## AgeMonths    0.002680   0.029728   0.090  0.92833    
## AoAE        -0.007632   0.015515  -0.492  0.62380    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.616 on 104 degrees of freedom
## Multiple R-squared:  0.07877,    Adjusted R-squared:  0.04334 
## F-statistic: 2.223 on 4 and 104 DF,  p-value: 0.07151

anova(plan_all_base, plan_all_AoAE)

## Analysis of Variance Table
## 
## Model 1: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths
## Model 2: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
##   Res.Df    RSS Df Sum of Sq     F Pr(>F)
## 1    105 1362.7                          
## 2    104 1359.5  1    3.1635 0.242 0.6238

plan_all_AoAE_AoLE <- lm(data=BRIEF_AgeOf, formula = Plan.Organize_RawScore~SES..3.66.+Sex+AgeMonths+AoAE+Age.of.Language.Exposure..mo.)
summary(plan_all_AoAE_AoLE)

## 
## Call:
## lm(formula = Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + 
##     AoAE + Age.of.Language.Exposure..mo., data = BRIEF_AgeOf)
## 
## Residuals:
##     Min      1Q  Median      3Q     Max 
## -6.3537 -2.3268 -0.6309  2.5441  8.6925 
## 
## Coefficients:
##                               Estimate Std. Error t value Pr(>|t|)    
## (Intercept)                   19.26342    2.13665   9.016 1.14e-14 ***
## SES..3.66.                    -0.06274    0.02622  -2.393  0.01853 *  
## SexMale                       -0.31561    0.68005  -0.464  0.64355    
## AgeMonths                     -0.02778    0.03029  -0.917  0.36136    
## AoAE                          -0.02023    0.01549  -1.306  0.19443    
## Age.of.Language.Exposure..mo.  0.06478    0.02122   3.054  0.00288 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Residual standard error: 3.479 on 103 degrees of freedom
## Multiple R-squared:  0.1552, Adjusted R-squared:  0.1142 
## F-statistic: 3.786 on 5 and 103 DF,  p-value: 0.003455

anova(plan_all_AoAE, plan_all_AoAE_AoLE)

## Analysis of Variance Table
## 
## Model 1: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE
## Model 2: Plan.Organize_RawScore ~ SES..3.66. + Sex + AgeMonths + AoAE + 
##     Age.of.Language.Exposure..mo.
##   Res.Df    RSS Df Sum of Sq      F   Pr(>F)   
## 1    104 1359.5                                
## 2    103 1246.7  1    112.86 9.3248 0.002877 **
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Findings: For all subscales, adding Age of Auditory Exposure to the model with demographic variables did not improve model fit (and Age of Auditory Exposure was not a significant predictor of any subsale), but subsequently adding Age of Language exposure to the model with Age of Auditory Exposure and demographic variables did significantly improve model fit (and Age of Language Exposure was a significant predictor of BRIEF-P scores in almost all models).

Check 2: Age-appropriate vs. non-age-appropriate participants

29 participants in the models included in this paper were outside the normed age range for the BRIEF-P (this is why we used raw scores in models above)

Possible issue: If this “non-age-appropriate” group of participants is largely from the Early ASL group, and/or has higher SES, our finding that Early exposure to language is the driving factor in EF development may not be valid

Approach: Checking the Age, SES, and Language Group of participants who were “age-appropriate” for the BRIEF-P compared to those who were not

table1::label(BRIEF_AgeOf$AgeMonths) <- "Age (Months)"
table1::label(BRIEF_AgeOf$SES..3.66.) <- "SES"
table1::label(BRIEF_AgeOf$LanguageGroup) <- "Language Group"


table1::table1(~AgeMonths + SES..3.66. + LanguageGroup | Age.Appropriate., data = BRIEF_AgeOf, overall=T)

	no (N=19)	yes (N=90)	TRUE (N=109)
Age (Months)
Mean (SD)	78.7 (5.85)	55.1 (9.93)	59.2 (13.0)
Median [Min, Max]	77.0 [72.0, 91.0]	56.0 [37.0, 71.0]	59.0 [37.0, 91.0]
SES
Mean (SD)	47.7 (20.3)	51.9 (11.6)	51.2 (13.5)
Median [Min, Max]	58.5 [8.00, 66.0]	54.0 [15.0, 66.0]	54.5 [8.00, 66.0]
Language Group
Typically Hearing	2 (10.5%)	44 (48.9%)	46 (42.2%)
Early ASL	4 (21.1%)	14 (15.6%)	18 (16.5%)
Later English	3 (15.8%)	20 (22.2%)	23 (21.1%)
Later ASL	10 (52.6%)	12 (13.3%)	22 (20.2%)

Findings:

Overall, Later-exposed groups make up 70% of the non-age-appropriate group.
ASL Later kids make up 48% of the non age-appropriate sample, 14.5% of the age-appropriate sample, and 21.6% of the overall sample
English later kids make up 20.7% of the non age-appropriate sample, 26.4% of the age-appropriate sample, and 25.2% of the overall sample

Participants in the Later groups, specifically the Later ASL group, are over-represented in not having received age-appropriate BRIEF. This should not be a concern for our findings, because if the behaviors surveyed in the BRIEF-P were not appropriate for older kids, then we would not expect the older Later ASL group to have elevated BRIEF scores relative to early-exposed kids–but they do.

Approach Pt 2: Re-run Chi-square with only age-appropriate subset to confirm the non-difference isn’t driven by older Early ASL participants Create dataframe with only age-appropriate participants exposed to language early

BRIEF_early_AgeApp <- subset(BRIEF_AgeApp, BRIEF_AgeApp$Language_Timing=="Early")

BRIEF Scores & Welch two sample t-tests for two “Early” groups within Age-appropriate subset

table1::label(BRIEF_early_AgeApp$GEC_RawScore) <- "Global Executive Composite"
table1::label(BRIEF_early_AgeApp$Inhibit_RawScore) <- "Inhibition"
table1::label(BRIEF_early_AgeApp$Shift_RawScore) <- "Shift"
table1::label(BRIEF_early_AgeApp$Emotional.Control_RawScore) <- "Emotional Control"
table1::label(BRIEF_early_AgeApp$Working.Memory_RawScore) <- "Working Memory"
table1::label(BRIEF_early_AgeApp$Plan.Organize_RawScore) <- "Plan/Organize"

table1(~GEC_RawScore + Inhibit_RawScore + Shift_RawScore + Emotional.Control_RawScore + Working.Memory_RawScore + Plan.Organize_RawScore | Language_Modality, data = BRIEF_early_AgeApp, overall=F)

	English (N=44)	ASL (N=20)
Global Executive Composite
Mean (SD)	88.4 (18.9)	91.4 (15.9)
Median [Min, Max]	88.0 [63.0, 155]	91.0 [63.0, 132]
Inhibition
Mean (SD)	22.8 (6.00)	24.3 (5.97)
Median [Min, Max]	23.0 [16.0, 46.0]	22.5 [16.0, 37.0]
Shift
Mean (SD)	13.2 (3.11)	14.0 (3.18)
Median [Min, Max]	12.0 [10.0, 24.0]	13.0 [10.0, 21.0]
Emotional Control
Mean (SD)	15.1 (3.69)	15.2 (3.14)
Median [Min, Max]	14.5 [10.0, 24.0]	14.0 [10.0, 21.0]
Working Memory
Mean (SD)	22.9 (6.72)	23.6 (5.00)
Median [Min, Max]	21.0 [17.0, 48.0]	23.0 [17.0, 37.0]
Plan/Organize
Mean (SD)	14.4 (3.44)	14.3 (3.05)
Median [Min, Max]	14.0 [10.0, 23.0]	14.0 [10.0, 20.0]

t.test(BRIEF_early_AgeApp$GEC_RawScore~BRIEF_early_AgeApp$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early_AgeApp$GEC_RawScore by BRIEF_early_AgeApp$Language_Modality
## t = -0.65032, df = 43.201, p-value = 0.5189
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -12.152931   6.225658
## sample estimates:
## mean in group English     mean in group ASL 
##              88.38636              91.35000

t.test(BRIEF_early_AgeApp$Inhibit_RawScore~BRIEF_early_AgeApp$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early_AgeApp$Inhibit_RawScore by BRIEF_early_AgeApp$Language_Modality
## t = -0.90549, df = 37.006, p-value = 0.3711
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -4.724033  1.805851
## sample estimates:
## mean in group English     mean in group ASL 
##              22.84091              24.30000

t.test(BRIEF_early_AgeApp$Shift_RawScore~BRIEF_early_AgeApp$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early_AgeApp$Shift_RawScore by BRIEF_early_AgeApp$Language_Modality
## t = -0.98755, df = 36.108, p-value = 0.3299
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -2.567680  0.885862
## sample estimates:
## mean in group English     mean in group ASL 
##              13.15909              14.00000

t.test(BRIEF_early_AgeApp$Emotional.Control_RawScore~BRIEF_early_AgeApp$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early_AgeApp$Emotional.Control_RawScore by BRIEF_early_AgeApp$Language_Modality
## t = -0.096402, df = 42.922, p-value = 0.9237
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.893154  1.720427
## sample estimates:
## mean in group English     mean in group ASL 
##              15.11364              15.20000

t.test(BRIEF_early_AgeApp$Working.Memory_RawScore~BRIEF_early_AgeApp$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early_AgeApp$Working.Memory_RawScore by BRIEF_early_AgeApp$Language_Modality
## t = -0.45497, df = 48.543, p-value = 0.6512
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -3.718724  2.345996
## sample estimates:
## mean in group English     mean in group ASL 
##              22.86364              23.55000

t.test(BRIEF_early_AgeApp$Plan.Organize_RawScore~BRIEF_early_AgeApp$Language_Modality)

## 
##  Welch Two Sample t-test
## 
## data:  BRIEF_early_AgeApp$Plan.Organize_RawScore by BRIEF_early_AgeApp$Language_Modality
## t = 0.12746, df = 41.286, p-value = 0.8992
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  -1.619017  1.837199
## sample estimates:
## mean in group English     mean in group ASL 
##              14.40909              14.30000

Supplementary Materials

Table 1: Comparing Global Executive Composite Models

stargazer(overall_all_base, overall_all_AoLE, overall_all_AoLE_AoAE, overall_all_AoLE_LoAE, overall_all_AoLE_Modality,overall_all_AoLE_HearStat, type= "html", title = "Linear Regression Results", align=TRUE, dep.var.labels=c("Global Executive Composite Raw Scores"), covariate.labels=c("Socioeconomic Status (SES)", "Sex (Male)", "Age (Months)", "Age of Exposure to Language (Months)", "Age of Auditory Exposure (Months)", "Length of Auditory Experience (Months)", "Language Modality (ASL)", "Hearing Status (Deaf)", "Hearing Status (HoH)"), out="BRIEF_supp_table1.html")

Table 2: Comparing Inhibition Models

stargazer(inhibition_all_base, inhibition_all_AoLE, inhibition_all_AoLE_AoAE, inhibition_all_AoLE_LoAE, inhibition_all_AoLE_Modality, inhibition_all_AoLE_HearStat, type= "html", title = "Linear Regression Results", align=TRUE, dep.var.labels=c("Inhibition Raw Scores"), covariate.labels=c("Socioeconomic Status (SES)", "Sex (Male)", "Age (Months)", "Age of Exposure to Language (Months)", "Age of Auditory Exposure (Months)", "Length of Auditory Experience (Months)", "Language Modality (ASL)", "Hearing Status (Deaf)", "Hearing Status (HoH)"), out="BRIEF_supp_table2.html")

Table 3: Comparing Shift Models.

stargazer(shift_all_base, shift_all_AoLE, shift_all_AoLE_AoAE, shift_all_AoLE_LoAE, shift_all_AoLE_Modality, shift_all_AoLE_HearStat, type= "html", title = "Linear Regression Results", align=TRUE, dep.var.labels=c("Shift Raw Scores"), covariate.labels=c("Socioeconomic Status (SES)", "Sex (Male)", "Age (Months)", "Age of Exposure to Language (Months)", "Age of Auditory Exposure (Months)", "Length of Auditory Experience (Months)", "Language Modality (ASL)", "Hearing Status (Deaf)", "Hearing Status (HoH)"), out="BRIEF_supp_table3.html")

Table 4: Comparing Emotional Control Models

stargazer(emotctrl_all_base, emotctrl_all_AoLE, emotctrl_all_AoLE_AoAE, emotctrl_all_AoLE_LoAE, emotctrl_all_AoLE_Modality, emotctrl_all_AoLE_HearStat, type= "html", title = "Linear Regression Results", align=TRUE, dep.var.labels=c("Emotional Control Raw Scores"), covariate.labels=c("Socioeconomic Status (SES)", "Sex (Male)", "Age (Months)", "Age of Exposure to Language (Months)", "Age of Auditory Exposure (Months)", "Length of Auditory Experience (Months)", "Language Modality (ASL)", "Hearing Status (Deaf)", "Hearing Status (HoH)"), out="BRIEF_supp_table4.html")

Table 5: Comparing Working Memory Models

stargazer(wm_all_base, wm_all_AoLE, wm_all_AoLE_AoAE, wm_all_AoLE_LoAE, wm_all_AoLE_Modality,wm_all_AoLE_HearStat, type= "html", title = "Linear Regression Results", align=TRUE, dep.var.labels=c("Working Memory Raw Scores"), covariate.labels=c("Socioeconomic Status (SES)", "Sex (Male)", "Age (Months)", "Age of Exposure to Language (Months)", "Age of Auditory Exposure (Months)", "Length of Auditory Experience (Months)", "Language Modality (ASL)", "Hearing Status (Deaf)", "Hearing Status (HoH)"), out="BRIEF_supp_table5.html")

Table 6: Comparing Planning/Organization Models

stargazer(plan_all_base, plan_all_AoLE, plan_all_AoLE_AoAE, plan_all_AoLE_LoAE, plan_all_AoLE_Modality, plan_all_AoLE_HearStat, type= "html", title = "Linear Regression Results", align=TRUE, dep.var.labels=c("Planning/Organization Raw Scores"), covariate.labels=c("Socioeconomic Status (SES)", "Sex (Male)", "Age (Months)", "Age of Exposure to Language (Months)", "Age of Auditory Exposure (Months)", "Length of Auditory Experience (Months)", "Language Modality (ASL)", "Hearing Status (Deaf)", "Hearing Status (HoH)"), out="BRIEF_supp_table6.html")

Table 7: Demographic information for the subset of participants who were in the appropriate age range for the BRIEF-P

table1::label(BRIEF_AgeApp$AgeMonths) <- "Age (Months)"
table1::label(BRIEF_AgeApp$Sex) <- "Sex"
table1::label(BRIEF_AgeApp$Race_recoded) <- "Race"
table1::label(BRIEF_AgeApp$Ethnicity_recoded) <- "Ethnicity"
table1::label(BRIEF_AgeApp$SES..3.66.) <- "SES"
 
table1::table1(~AgeMonths + Sex + Race_recoded + Ethnicity_recoded + SES..3.66. | LanguageGroup, data = BRIEF_AgeApp, overall=F)

	Typically Hearing (N=44)	Early ASL (N=20)	Later English (N=20)	Later ASL (N=14)
Age (Months)
Mean (SD)	53.7 (9.80)	54.5 (8.98)	58.3 (9.67)	56.8 (11.3)
Median [Min, Max]	54.0 [37.0, 71.0]	54.5 [41.0, 71.0]	60.0 [37.0, 71.0]	60.0 [37.0, 71.0]
Sex
Female	23 (52.3%)	11 (55.0%)	10 (50.0%)	7 (50.0%)
Male	21 (47.7%)	9 (45.0%)	10 (50.0%)	7 (50.0%)
Race
Asian	0 (0%)	0 (0%)	0 (0%)	1 (7.1%)
Black or African American	0 (0%)	0 (0%)	1 (5.0%)	0 (0%)
White	41 (93.2%)	18 (90.0%)	17 (85.0%)	11 (78.6%)
More than one	3 (6.8%)	1 (5.0%)	2 (10.0%)	1 (7.1%)
Other/Missing	0 (0%)	1 (5.0%)	0 (0%)	1 (7.1%)
Ethnicity
Hispanic	3 (6.8%)	0 (0%)	1 (5.0%)	2 (14.3%)
Non-Hispanic	40 (90.9%)	13 (65.0%)	15 (75.0%)	10 (71.4%)
Prefer not to answer	0 (0%)	1 (5.0%)	0 (0%)	0 (0%)
Missing	1 (2.3%)	6 (30.0%)	4 (20.0%)	2 (14.3%)
SES
Mean (SD)	55.5 (8.83)	47.0 (16.0)	49.6 (11.3)	46.7 (14.2)
Median [Min, Max]	56.0 [21.5, 66.0]	53.8 [19.0, 66.0]	50.5 [15.0, 66.0]	50.8 [9.00, 62.0]

Table 8. Relative Risk of Clinically-Elevated (T-score ≥ 60) Executive Functioning Deficits in DHH Participants

Risk Ratios (Relative Risk) and 95% confidence intervals for Global Executive Composite T-scores > 60

BRIEF_AgeApp_GEC_over60 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Global Executive Composite T-score < 60 (n)" = sum(GEC_Tscore < 60), "Global Executive Composite T-score > 60 (n)" = sum(GEC_Tscore >= 60))

BRIEF_AgeApp_GEC_over60 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Global Executive Composite T-score < 60 (n)	Global Executive Composite T-score > 60 (n)
Typically Hearing	39	5
Early ASL	17	3
Later English	14	6
Later ASL	10	4

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
GEC_60rr <- matrix(c(39,5,17,3, 14, 6, 10, 4), 4, 2, byrow=TRUE)
dimnames(GEC_60rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "GEC T-score" = c("Not Elevated", "Elevated"))
riskratio.small(GEC_60rr)

## $data
##                    GEC T-score
## Group               Not Elevated Elevated Total
##   Typically Hearing           39        5    44
##   Early ASL                   17        3    20
##   Later English               14        6    20
##   Later ASL                   10        4    14
##   Total                       80       18    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate     lower    upper
##   Typically Hearing 1.000000        NA       NA
##   Early ASL         1.125000 0.2974811 4.254472
##   Later English     2.250000 0.7774878 6.511356
##   Later ASL         2.142857 0.6656236 6.898548
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL         0.68785362   0.69697927 0.68348090
##   Later English     0.08984932   0.08354265 0.06699369
##   Later ASL         0.16123264   0.19823228 0.12141705
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

#DHH groups not significantly more likely to have elevated GEC T-scores relative to TH group, but the two later groups are close

Inhibit T-scores > 60 Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_Inhibit_over60 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Inhibit T-score < 60 (n)" = sum(Inhibit_Tscore < 60), "Inhibit T-score > 60 (n)" = sum(Inhibit_Tscore >= 60))

BRIEF_AgeApp_Inhibit_over60 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Inhibit T-score < 60 (n)	Inhibit T-score > 60 (n)
Typically Hearing	39	5
Early ASL	16	4
Later English	17	3
Later ASL	10	4

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
Inhibit_60rr <- matrix(c(39,5,16,4, 17, 3, 10, 4), 4, 2, byrow=TRUE)
dimnames(Inhibit_60rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Inhibit T-score" = c("Not Elevated", "Elevated"))
riskratio.small(Inhibit_60rr)

## $data
##                    Inhibit T-score
## Group               Not Elevated Elevated Total
##   Typically Hearing           39        5    44
##   Early ASL                   16        4    20
##   Later English               17        3    20
##   Later ASL                   10        4    14
##   Total                       82       16    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate     lower    upper
##   Typically Hearing 1.000000        NA       NA
##   Early ASL         1.500000 0.4500510 4.999433
##   Later English     1.125000 0.2974811 4.254472
##   Later ASL         2.142857 0.6656236 6.898548
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.3869141    0.4434313  0.3569397
##   Later English      0.6878536    0.6969793  0.6834809
##   Later ASL          0.1612326    0.1982323  0.1214171
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

#DHH groups *not* significantly more likely to have elevated Inhibit T-scores relative to TH group, but the Later ASL group close

Shift T-scores > 60 Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_Shift_over60 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Shift T-score < 60 (n)" = sum(Shift_Tscore < 60), "Shift T-score > 60 (n)" = sum(Shift_Tscore >= 60))

BRIEF_AgeApp_Shift_over60 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Shift T-score < 60 (n)	Shift T-score > 60 (n)
Typically Hearing	42	2
Early ASL	18	2
Later English	18	2
Later ASL	10	4

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
Shift_60rr <- matrix(c(42,2,18,2, 18, 2, 10, 4), 4, 2, byrow=TRUE)
dimnames(Shift_60rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Shift T-score" = c("Not Elevated", "Elevated"))
riskratio.small(Shift_60rr)

## $data
##                    Shift T-score
## Group               Not Elevated Elevated Total
##   Typically Hearing           42        2    44
##   Early ASL                   18        2    20
##   Later English               18        2    20
##   Later ASL                   10        4    14
##   Total                       88       10    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate     lower    upper
##   Typically Hearing 1.000000        NA       NA
##   Early ASL         1.500000 0.2272096  9.90275
##   Later English     1.500000 0.2272096  9.90275
##   Later ASL         4.285714 0.8763638 20.95859
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL         0.45602919   0.58311299  0.4033953
##   Later English     0.45602919   0.58311299  0.4033953
##   Later ASL         0.02789668   0.02564615  0.0101395
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

#Later ASL group significantly more likely to have elevated Shift T-scores relative to TH group (3.75 times more likely) according to p-values (even though risk ratio CI does include 1)

Emotional Control T-scores > 60 Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_EmoCtrl_over60 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Emotional Control T-score < 60 (n)" = sum(Emotional.Control_Tscore < 60), "Emotional Control T-score > 60 (n)" = sum(Emotional.Control_Tscore >= 60))

BRIEF_AgeApp_EmoCtrl_over60 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Emotional Control T-score < 60 (n)	Emotional Control T-score > 60 (n)
Typically Hearing	33	11
Early ASL	18	2
Later English	18	2
Later ASL	11	3

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
EmoCtrl_60rr <- matrix(c(33,11, 18,2, 18,2, 11,3), 4, 2, byrow=TRUE)
dimnames(EmoCtrl_60rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Emotional Control T-score" = c("Not Elevated", "Elevated"))
riskratio.small(EmoCtrl_60rr)

## $data
##                    Emotional Control T-score
## Group               Not Elevated Elevated Total
##   Typically Hearing           33       11    44
##   Early ASL                   18        2    20
##   Later English               18        2    20
##   Later ASL                   11        3    14
##   Total                       80       18    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group                estimate      lower    upper
##   Typically Hearing 1.0000000         NA       NA
##   Early ASL         0.3750000 0.09147347 1.537331
##   Later English     0.3750000 0.09147347 1.537331
##   Later ASL         0.8035714 0.26060473 2.477802
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.1830447     0.200987  0.1668176
##   Later English      0.1830447     0.200987  0.1668176
##   Later ASL          0.8196528     1.000000  0.7856283
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

#Later English borderline *less* likely than TH to have elevated Emotional Control T-scores (but there do seem to be an unexpectedly high number of TH kids in the "elevated" score range, and Later English group only on average 4 months older than TH kids)

Working Memory T-scores > 60 Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_WorkMem_over60 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Working Memory T-score < 60 (n)" = sum(Working.Memory_Tscore < 60), "Working Memory T-score > 60 (n)" = sum(Working.Memory_Tscore >= 60))

BRIEF_AgeApp_WorkMem_over60 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Working Memory T-score < 60 (n)	Working Memory T-score > 60 (n)
Typically Hearing	38	6
Early ASL	16	4
Later English	15	5
Later ASL	8	6

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
WorkMem_60rr <- matrix(c(38,6,16,4,15,5,8,6), 4, 2, byrow=TRUE)
dimnames(WorkMem_60rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Working Memory T-score" = c("Not Elevated", "Elevated"))
riskratio.small(WorkMem_60rr)

## $data
##                    Working Memory T-score
## Group               Not Elevated Elevated Total
##   Typically Hearing           38        6    44
##   Early ASL                   16        4    20
##   Later English               15        5    20
##   Later ASL                    8        6    14
##   Total                       77       21    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate     lower    upper
##   Typically Hearing 1.000000        NA       NA
##   Early ASL         1.285714 0.4073295 4.058290
##   Later English     1.607143 0.5553484 4.650969
##   Later ASL         2.755102 1.0564549 7.184961
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL         0.53280902   0.71159622 0.51576333
##   Later English     0.29363909   0.29746086 0.26403699
##   Later ASL         0.03310511   0.05208747 0.01873012
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

#Later ASL significantly more likely to have elevated Working Memory T-scores relative to TH group

Plan/Organize T-scores > 60 Risk Ratios and 95% Confidence Intervals

BRIEF_AgeApp_PlanOrg_over60 <- BRIEF_AgeApp %>% 
    group_by(LanguageGroup) %>% 
    summarise("Plan/Organize T-score < 60 (n)" = sum(Plan.Organize_Tscore < 60), "Plan/Organize T-score > 60 (n)" = sum(Plan.Organize_Tscore >= 60))

BRIEF_AgeApp_PlanOrg_over60 %>%
    kable() %>%  
    kable_styling(bootstrap_options = "striped")

LanguageGroup	Plan/Organize T-score < 60 (n)	Plan/Organize T-score > 60 (n)
Typically Hearing	37	7
Early ASL	17	3
Later English	14	6
Later ASL	10	4

#The numbers in this matrix are static generated from the code above. If the numbers in the matrix change, the code also needs to be altered
PlanOrg_60rr <- matrix(c(37,7,17,3, 14, 6, 10, 4), 4, 2, byrow=TRUE)
dimnames(PlanOrg_60rr) <- list("Group" = c("Typically Hearing", "Early ASL", "Later English", "Later ASL"), "Plan/Organize T-score" = c("Not Elevated", "Elevated"))
riskratio.small(PlanOrg_60rr)

## $data
##                    Plan/Organize T-score
## Group               Not Elevated Elevated Total
##   Typically Hearing           37        7    44
##   Early ASL                   17        3    20
##   Later English               14        6    20
##   Later ASL                   10        4    14
##   Total                       78       20    98
## 
## $measure
##                    risk ratio with 95% C.I.
## Group               estimate     lower    upper
##   Typically Hearing 1.000000        NA       NA
##   Early ASL         0.843750 0.2429636 2.930127
##   Later English     1.687500 0.6501810 4.379790
##   Later ASL         1.607143 0.5506079 4.691012
## 
## $p.value
##                    two-sided
## Group               midp.exact fisher.exact chi.square
##   Typically Hearing         NA           NA         NA
##   Early ASL          0.9529974    1.0000000  0.9260295
##   Later English      0.2210706    0.3140510  0.1940420
##   Later ASL          0.3257076    0.4333345  0.2925137
## 
## $correction
## [1] FALSE
## 
## attr(,"method")
## [1] "small sample-adjusted UMLE & normal approx (Wald) CI"

#DHH groups *not* significantly more likely to have elevated Inhibit T-scores relative to TH group

Language Not Auditory Experience is Related to Parent-Reported Executive Functioning in Preschool-Aged Deaf and Hard of Hearing Children

Corina Coodwin, Ph.D.

Emily Carrigan, Ph.D. (code author)

Kristin Walker (code co-author)

Marie Coppola, Ph.D

Load Packages

Setup Primary Dataframe

Demographic Information

Demographics for the Entire Sample

Table 1.

Comparing SES across Language Groups

Comparing Age across Language Groups

Results

Question 1: Do DHH children with access to sign language but not auditory input from birth exhibit executive functioning skills comparable to their typically-hearing peers?

Demographics for the Early-exposed subset

Comparing demographics by Language Modality (within early-exposed subset)

Table 3.

Question 2: Do age of access to auditory input and age of language exposure significantly predict executive functioning?

Demographics for the Question 2 subset

Comparing demographics by Language group (within Question 2 subset)

Table 2.

Comparing Age of Auditory Exposure across Language Groups

Comparing Age of Language Exposure across Language Groups

Checking correlations between continuous (numerical) predictors

Global Executive Composite Models

Inhibition Models

Shift Models

Emotional Control Models

Working Memory Models

Planning/Organization Models

Table 4.

Question 3: Do children with later or degraded access to auditory and/or language input exhibit clinically-significant deficits in executive functioning skills relative to typically-hearing peers?

Demographics for the Question 3 subset

Comparing demographics by Language group (within Question 3 subset)

Figure 1.

Table 5 data & analyses

Risk Ratios (Relative Risk) and 95% confidence intervals for Global Executive Composite T-scores

Inhibit Risk Ratios and 95% Confidence Intervals

Shift Risk Ratios and 95% Confidence Intervals

Emotional Control Risk Ratios and 95% Confidence Intervals

Working Memory Risk Ratios and 95% Confidence Intervals

Plan/Organize Risk Ratios and 95% Confidence Intervals

Additional checks

Check 1: Comparing stepwise models with AoAE added first

Check 2: Age-appropriate vs. non-age-appropriate participants

Supplementary Materials

Table 1: Comparing Global Executive Composite Models

Table 2: Comparing Inhibition Models

Table 3: Comparing Shift Models.

Table 4: Comparing Emotional Control Models

Table 5: Comparing Working Memory Models

Table 6: Comparing Planning/Organization Models

Table 7: Demographic information for the subset of participants who were in the appropriate age range for the BRIEF-P

Table 8. Relative Risk of Clinically-Elevated (T-score ≥ 60) Executive Functioning Deficits in DHH Participants

Risk Ratios (Relative Risk) and 95% confidence intervals for Global Executive Composite T-scores > 60

Inhibit T-scores > 60 Risk Ratios and 95% Confidence Intervals

Shift T-scores > 60 Risk Ratios and 95% Confidence Intervals

Emotional Control T-scores > 60 Risk Ratios and 95% Confidence Intervals

Working Memory T-scores > 60 Risk Ratios and 95% Confidence Intervals

Plan/Organize T-scores > 60 Risk Ratios and 95% Confidence Intervals